Ask the Scholar

Document scope · 1 page
doc
Scholar
Ask about this object, its catalog metadata, its source description, or the page inventory. For page-specific OCR and visual context, open one of the page chats.

Scholar Source Context

Document identity
localId
34428929
label
Medicaid/Medicare [4]
core
doc
dtoType
document
pageCount
1
Source metadata
Source extras
naId
34428929
levelOfDescription
fileUnit
otherTitles
42-t-7422560-20140536S-032-010-2016
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
543ebabe8ec17313
ocrText
The Development of Capitation Rates under Medicaid Managed Care Programs: A Pilot Study Volume 1: Summary and Analysis of Findings HR Health Systems Research, Inc. Prepared for THE HENRY KAISER FAMILY FOUNDATION The Henry J. Kaiser Family Foundation November 1997 The Henry J. Kaiser Family Foundation, based in Menlo Park, California, is a non-profit, independent national health care philanthropy and is not associated with Kaiser Permanente or Kaiser Industries. The Foundation's work is focused on four main areas: health policy, reproductive health, and HIV/AIDS policy in the United States, and health and development in South Africa. The Development of Capitation Rates under Medicaid Managed Care Programs: A Pilot Study Volume I: Summary and Analysis of Findings November 1997 Prepared for: The Henry J. Kaiser Family Foundation Prepared by: Renee Schwalberg, MPH Health Systems Research, Inc. 2021 L Street, N.W. Suite 400 Washington, D.C. 20036 Table of Contents Executive Summary ii Chapter I. Introduction and Overview 1 Chapter II. Review of Medicaid Managed Care Initiatives and Rate Development Methodologies 4 A. Medicaid Managed Care Program Design 4 B. Rate Development 5 C. Technical Criteria 15 D. Actuarial Methods 16 E. Negotiation Methods 28 F. Risk and Profit Sharing 35 G. Enrollment Provisions Related to Capitation Rates 36 H. Quality Assurance Provisions Related to Capitation Rates 39 Chapter III. Conclusions and Future Directions 42 Attachment A: Interview Protocol Health Systems Research, Inc. Table of Contents Page i Executive Summary I. Introduction and Overview State Medicaid agencies are increasingly turning to capitated managed care systems to improve access to appropriate care and control the costs of their Medicaid programs. This pilot study explores several aspects of states' systems for contracting with managed care plans, focusing on the population of AFDC and related eligibles in 15 study states. The study questions included the methods used by states to develop capitation rates for AFDC and related eligibles, to what extent and how rates are negotiated with managed care plans, the benefits provided for these rates, and the quality assurance mechanisms that are tied to capitation rates. This pilot study discussed these questions with officials of 15 study states, focusing on Medicaid managed care systems for AFDC and related populations, including AFDC cash recipients, AFDC non-cash eligibles, and groups of pregnant women and children covered through the Medicaid expansions of the late 1980s. The 15 study states were chosen to represent a range of managed care penetration rates and years of experience with capitated programs for the Medicaid population, as well as broad geographic representation. The study states are listed in Table 1 on the next page. The study methodology included detailed interviews with Medicaid officials using a standard protocol, followed by a review of a range of documents from each state, including Requests for Proposals or Applications, standard contracts, and actuarial reports. This volume presents a summary and analysis of the results of the study. A summary of the individual states' actuarial methods and contracting processes is presented in Volume II of this report. Health Systems Research, Inc. Executive Summary Page ii Table 1. Study States Year Risk-Based Medicaid Penetration Rate Managed Care Introduced as of 1995* Before 1990 1990 and After >50% Arizona Oregon California Tennessee Rhode Island Washington 10%-50% Florida Connecticut Minnesota Delaware Missouri Massachusetts New York <10% Georgia Texas *Percentage of total Medicaid eligibles enrolled in capitated arrangements. Source: Health Systems Research, Inc. II. Review of Medicaid Managed Care Initiatives and Rate Development Methodologies A. Medicaid Managed Care Program Design The 15 study states included a range of types of Medicaid managed care programs, as described below: Mandatory VS. Voluntary Systems. Eleven of the study states require AFDC and related Medicaid eligibles to enroll in capitated systems, while enrollment by these populations is an option in the remaining four. Waivers. Six of the study states are using section 1115(a) research and demonstration waivers to operate capitated managed care programs, and eight are using section 1915(b) waiver authority to operate capitated programs. One study state had no waivers. Eligibility Categories Covered. All of the study states enroll AFDC and related groups, including children in the eligibility groups mandated by the Omnibus Health Systems Research, Inc. Executive Summary Page iii Budget Reconciliation Act (OBRA) of 1989, in capitated arrangements. Two states exempt pregnant women who do not receive AFDC (those in the SOBRA eligibility group) from capitated systems. Geographic Coverage. Nine of the study states operate statewide Medicaid managed care programs, while six operate capitated arrangements only in designated counties. Benefits. All of the study states include basic required Medicaid benefits, including physician services, inpatient and outpatient hospital care, primary and preventive care, and the Early and Periodic Screening, Diagnosis, and Treatment program (EPSDT) in their capitated systems. Services that may be excluded from plans' benefit packages include specialty services for children enrolled in the state's Title V Children with Special Health Care Needs program; behavioral health services; Part H Early Intervention services; and organ transplants. B. Rate Development The general methods used by the study states to contract with plans and establish capitation rates can be classified according to two criteria: how the state uses technical criteria to select plans for contracts, and how capitation rates are developed. Within each of these areas, the study states may again be grouped into two categories. For both its technical and price criteria, a state may either develop a set of standards that all contracting plans must meet, or may score plans based on their technical and cost proposals. These two distinctions may be combined to produce four types of contracting methodologies: No Price Competition. Six study states award contracts to all plans that meet the state's technical standards and are willing to accept the capitation rates determined by the state. Non-price Competition. In this model, the state determines the number of plans needed in each geographic area and plans compete on technical standards to win these contracts. Two study states select plans based on competition only on technical criteria; the plan or plans selected must accept the rates developed by the state. Plans Meet Technical Standards and Negotiate Rates. Two study states require plans to meet specific technical standards but do not score the plans on their technical proposals. Rather, the submission of an acceptable technical Health Systems Research, Inc. Executive Summary Page iv proposal is a prerequisite for the evaluation of the plan's cost proposal, which is used to negotiate a capitation rate. In these systems, there is no predetermined limit on the number of plans with which the state will contract; thus, the plans are not in direct competition with each other. Plans Scored on Technical and Cost Proposals. The remaining five states in the study use systems in which plans are selected based on the level of their proposed capitation rates as well as technical criteria. In these states, the number of contracts to be awarded is limited; thus, the plans are in direct competition with each other on both technical and cost criteria. The use of these methodologies in the 15 study states is illustrated in Table 2 below. Table 2. Classification of Contracting Methodologies Technical Standards Capitation Rates Plans Meet State Thresholds Plans Scored Plans Accept State's Connecticut California Rates Florida Delaware Georgia Minnesota Oregon Tennessee Plans Present Cost Massachusetts Arizona Proposals Rhode Island Missouri New York Texas Washington State officials reported a number of reasons for choosing one or another of these contracting methods. Under the first two of these models, states develop capitation rates directly, without requiring cost proposals or negotiation with plans. This method is seen as efficient, especially in states that were required to implement their managed care systems soon after receiving 1115(a) waivers. In addition, this is seen as an attractive option for states that are beginning to implement managed care systems, as state Medicaid agencies are accustomed to setting fee-for- service rates. Finally, state officials point out that, even in competitive bidding systems, the Health Systems Research, Inc. Executive Summary Page V state would need to develop a benchmark against which to compare plans' bids; therefore, a non-competitive system in which the state develops capitation rates seems like a logical first step, even if states plan to implement competitive bidding systems in the future. The major advantage reported for price competition is the potential it offers for cost savings. The desire to win a contract in a competitive bidding system presents a clear incentive for plans to present low bids. In addition, four states have designed algorithms for assigning those eligibles who do not voluntarily choose a managed care plan to favor those plans that offer the lowest capitation rates or receive the highest total point scores. In addition, in many states, competition itself is valued by state officials and legislators. However, a disadvantage of a competitive process is the potential for significant disruptions to occur in later years if a participating plan or plans lose their contracts and their members are forced to switch to other plans. C. Technical Criteria Each study state establishes technical criteria for managed care plans contracting with the state. Technical standards may be set that all plans must meet in order to receive contracts, technical criteria may be the only criteria on which plans compete, or scores on a cost proposal may be combined with scores on the technical proposal to develop a total score for each bidder. Examples of the types of technical criteria used in the study states include the quality, size, and comprehensiveness of the plan's provider network; the ability to meet outcome goals, such as improvements in infant mortality and low birth weight rates; commitment to internal quality improvement and ability to provide encounter data; the quality of program administration, including executive management and grievance procedures; the ability of the plan to perform administrative functions, including data collection and reporting; financial stability; cultural and linguistic requirements; and the ability to provide health education services. Health Systems Research, Inc. Executive Summary Page vi D. Actuarial Methods States' methods for establishing capitation rates can be divided into two broad categories: those that establish specific rates to be paid to all contracting plans, or those that require plans to present cost proposals, with final rates agreed upon through negotiation between plans and the state Medicaid agency. In both of these types of systems, state Medicaid agencies are required to develop Upper Payment Limits (UPLs) to guide the selection of and negotiation with plans and to assure that the rates established meet the fiscal guidelines required by their Medicaid waivers. Thus, all study states calculated either capitation rates or guidelines against which to evaluate plans' cost proposals. These rates and rate guidelines used by states are set using a variety of parameters, including enrollees' age and sex, eligibility category, and geographic location. Nearly all study states used age and sex groups to develop age- and sex-specific rates or to adjust final rates for the demographic distribution of each plan or geographic area. Twelve states set distinct rates for specific geographic areas within the state, to reflect differences in the cost of providing care in different areas of the state. Finally, nine states set distinct rates for specific eligibility groups. In five of these, all AFDC and related categories, including pregnant women and children in expansion categories, are grouped in one eligibility category for rate development purposes. In most states, the rates within each cell are calculated using databases of past years' fee-for- service claims. From these raw databases, states generally first subtract claims for population or services excluded from the Medicaid managed care program. The total remaining expenditures within each category are then divided by the number of enrollee months represented by the category to develop base per-member-per-month (PMPM) expenditures. In every study state, these base calculations are subject to a number of adjustments, including both increases and decreases in the rates, in the process of arriving at final rates or UPLs. Examples of adjustments used by the study states are presented below. Health Systems Research, Inc. Executive Summary Page vii 1. Increases Base PMPM rates may be increased by a variety of factors. Common factors used in the study states include the following: Trending rates forward to the contract year; An increase for the administrative costs to the managed care plan; and Claims completion factors. 2. Decreases Many states also apply a number of factors that reduce the base PMPM rates, including the following: A prescription drug rebate adjustment; An adjustment for third-party liability; An adjustment for voluntary selection (in states in which enrollment in capitated systems is voluntary); and Desired or expected savings from the use of managed care. E. Negotiation Methods Seven of the 15 study states require plans to present cost proposals, which are compared to the state's rate guidelines. These guidelines always include an upper limit, often based on the Upper Payment Limit required by 1915(b) waivers. They may also include explicit lower limits, which represent the lowest rate state officials feel would be adequate to support the provision of comprehensive, high-quality care. Both the upper and lower payment limits may or may not be shared with plans through the Request for Proposals. The study states' methods for sharing rate guidelines and negotiating rates with offering plans are summarized below. Explicit Upper and Lower Limits Shared with Plans. Two study states set explicit upper and lower limits and share them with plans. This approach both creates an atmosphere of openness in the relationship between the state and the plans, and increases the likelihood that the plans will bid rates that fall within the acceptable range. The clear disadvantage of sharing upper and lower limits Health Systems Research, Inc. Executive Summary Page viii with plans is that the state runs the risk of receiving only bids at the top of the range. States may mitigate this problem by developing a scoring system that rewards plans for offering low bids. Explicit Upper and Lower Limits Not Shared with Plans. To avoid offering plans an incentive to bid at the top of a range, three study states set explicit upper and lower payment limits but do not share them with plans. In these systems, the plans' cost proposals are compared to the state's acceptable range of rates and given a score; if plans bid above or below the range, the state develops a counter-offer based on a standard algorithm, and scores the plan based on the counter-offer. Explicit Upper Limit Shared with Plans. One study state sets an upper payment limit and shares it with plans as part of its RFP. Because the upper limit provides a degree of cost savings that is satisfactory to state officials, the risk that most bids would be close to the limit is not seen as a disadvantage. Explicit Upper Limit Not Shared with Plans. One study state sets only an upper payment limit that is not shared with plans. The state has historically struggled to contract with plans for rates below this limit, so state officials have not felt it necessary to set an explicit lower bound for capitation rates. F. Risk and Profit Sharing The most common mechanism for sharing risk with plans is through the availability of state- funded reinsurance. Of the study states, five offer state-funded reinsurance to plans; two additional states offered this reinsurance in the past, but have stopped, since plans preferred to buy such insurance on the private market. When plans do accept state-sponsored reinsurance, their capitation rates will be reduced by an amount actuarially equivalent to the value of the benefit. An additional seven states require plans to maintain private reinsurance. Four states reported that they have specific additional mechanisms for sharing risk or profits with plans; in two cases, these arrangements are provided to protect plans organized by traditional safety-net providers. G. Enrollment Provisions Related to Capitation Rates In states that require plans to compete for contracts, the enrollment of those eligibles who do not voluntarily choose a managed care plan may be used as an incentive for plans to present low bids. Plans may be given preference in enrollment of non-choosers based on price, with Health Systems Research, Inc. Executive Summary Page ix the plans bidding the lowest rates receiving a plurality of the auto-enrolled population; on quality, with the plan receiving the highest technical score receiving preference; or a combination of the two, with the plans with the highest total score receiving a plurality of assignees. All of these strategies are based on the assumption that those who fail to choose a plan are likely to be relatively healthy, and thus will not use high levels of services. Therefore, plans would be expected to compete, whether through price competition, enhanced levels of quality, or both, to receive preference in enrolling this population. Two states designed auto-enrollment mechanisms on a different assumption: that those who do not choose a plan represent a random sample of the eligible population. In these states, the auto-enrolled population is assigned to plans based on the enrollment distribution of those who do voluntarily choose a plan. H. Quality Assurance Provisions Related to Capitation Rates Each state's contract with managed care plans outlines a wide range of quality assurance and improvement efforts, including the use of practice guidelines, member satisfaction surveys, HEDIS measures, and internal quality improvement systems. Several states also require plans to submit monthly or quarterly encounter data. These data, in addition to their use in calculating quality indicators, may be used to assess the adequacy of current rates and to calculate rates or rate guidelines for future years. In addition, three states have developed mechanisms to tie some portion of capitation payments to the achievement of specific quality objectives, such as EPSDT screening rates, rates of other preventive health services, or overall compliance with contract requirements. III. Conclusion and Future Directions In developing their strategies for establishing capitation rates, states must balance the competing interests of controlling Medicaid program costs (and, ideally, achieving savings) and attracting plans to the managed care initiative. While most states reported that their rate- development methodologies allowed them to meet both goals successfully, several have had difficulty striking an appropriate balance. Health Systems Research, Inc. Executive Summary Page X Another overriding issue of concern in many states is the need to identify appropriate sources of data on which to base capitation rates. Most study states are currently using claims data from their fee-for-service Medicaid programs as the basis for their rate calculations. However, using fee-for-service claims as the basis for capitation rates may preserve inequities and inefficiencies found in the fee-for-service system. In addition, as capitation programs expand, new fee-for-service data will no longer be available, or will be based on geographic areas whose utilization patterns are not comparable to those in areas in which capitated managed care is being implemented. To address these concerns, several states hope to use encounter data submitted by managed care plans as the basis for future rates. Overall, the survey demonstrated the wide variety of systems currently in use to establish capitation rates and negotiate contracts in Medicaid managed care systems. No single methodology emerged as clearly the most successful, and officials of the study states reported a variety of advantages and disadvantages of each method used. Moreover, states are continuing to modify their approaches to Medicaid managed care contracting, by modifying the methods used to develop rates for AFDC and related populations as well as by developing methods for setting specific rates for special populations. Health Systems Research, Inc. Executive Summary Page xi CHAPTER I Introduction and Overview State Medicaid agencies are increasingly turning to capitated managed care systems to improve access to appropriate care and control the cost of their Medicaid programs. As of 1996, 38 states used capitated systems to serve at least some portion of their Medicaid-enrolled populations.¹ Although substantial research has described the program characteristics of these Medicaid managed care programs, few studies have analyzed the processes used to establish the capitation rates paid to managed care plans. This study, commissioned by the Kaiser Commission on the Future of Medicaid, examines the following questions: What actuarial methods are used by states to develop capitation rates for AFDC and related eligibles? To what extent and how are rates negotiated with managed care plans? What benefits are provided under these rates, and what types of quality assurance data are states collecting to monitor the use of Medicaid funds? This pilot study explored these questions by focusing on rates paid for the coverage of AFDC and related populations in 15 states. The study focused on groups that represent predominantly low-income women and children, including AFDC cash recipients, AFDC non-cash eligibles, and groups of pregnant women and children covered through the Medicaid expansions of the late 1980s. Medicaid eligibles in other categories, such as the aged, blind and disabled, Qualified Medicare Beneficiaries, and those enrolled in Medically Needy programs, were not included in the pilot study. 1 National Academy for State Health Policy, 1997. Health Systems Research, Inc. Chapter I Page 1 The 15 study states were chosen to include states with a range of managed care penetration rates and years of experience within the Medicaid population and to represent all regions of the country. The states chosen for the study are presented in Table 1 below. Table 1. Study States Year Risk-Based Medicaid Penetration Rate Managed Care Introduced as of 1995* Before 1990 1990 and After >50% Arizona Oregon California Tennessee Rhode Island Washington 10%-50% Florida Connecticut Minnesota Delaware Missouri Massachusetts New York <10% Georgia Texas *Percentage of total Medicaid eligibles enrolled in capitated arrangements. Source: Health Systems Research, Inc. With each of these states, a detailed interview using a standard protocol, was conducted with a representative of the state Medicaid agency who was closely involved in the rate determination and contracting processes (the protocol is included here as Attachment A). This interview explored a range of aspects of the contracting process, including actuarial methods, negotiation processes, and quality assurance requirements. (In cases in which a state had more than one capitated program for AFDC and related groups, information was gathered about all programs.) These interviews were conducted in late 1996 and early 1997; the information presented here therefore reflects the systems and processes that were in place at that time. In addition to the telephone interviews, HSR collected and reviewed a variety of public documents relating to each state's contracting process, including Requests for Proposals or Applications, standard Health Systems Research, Inc. Chapter I Page 2 contracts, and actuarial reports. For each state, a summary was developed that describes the state's actuarial methods and contracting process; these are included in Volume II of this report. The present volume includes an overall description and analysis of the study findings. The following section discusses the design of Medicaid managed care programs in the study states, the methods used to develop and negotiate rates, and provisions for enrollment, risk sharing, and quality assurance. The final section presents the conclusion and lessons learned by the study states about the process of developing capitation rates. Health Systems Research, Inc. Chapter I Page 3 CHAPTER II Review of Medicaid Managed Care Initiatives and Rate Development Methodologies A. Medicaid Managed Care Program Design The 15 study states included a range of types of Medicaid managed care programs, including those that require AFDC and related eligibles to enroll in capitated managed care plans and those that offer this as an option; those that are implemented statewide and those that currently cover only selected counties; and those that are operating under section 1915(b) freedom-of- choice waivers and those that are using section 1115(a) research and demonstration waivers. Table 2 presents the program characteristics of each state's major Medicaid managed care initiative; the major areas of variation are summarized briefly below. Mandatory VS. Voluntary Systems. All of the study states except Georgia and New York require that AFDC and related groups enroll in some form of managed care, either a capitated plan or a primary care case management (PCCM) model. Eleven of the study states require these eligibles to enroll in capitated systems, while enrollment in capitated systems by these populations is an option in the remaining four. Waivers. Six of the study states are using section 1115(a) research and demonstration waivers to operate capitated managed care programs, and eight are using section 1915(b) waiver authority to operate capitated programs. In several of these states, 1115(a) waiver applications were pending at the time of the interviews. One study state, Georgia, had no waivers. Eligibility Categories Covered. As mentioned earlier, all of the study states enroll AFDC and related groups, including children in the eligibility groups mandated by the Omnibus Budget Reconciliation Act (OBRA) of 1989, in managed care arrangements. Two states, Georgia and Florida, exempt pregnant women who do not receive AFDC (those in the SOBRA eligibility group) from capitated systems; this will be discussed in detail below. Health Systems Research, Inc. Chapter II Page 4 In addition to AFDC and related populations, several states, including Tennessee and Oregon, include those eligible for Supplemental Security Income (SSI) and those in Medically Needy categories in their capitated programs. In addition, five study states have used their 1115(a) waiver authority to expand eligibility to populations of uninsured adults and children not otherwise eligible for Medicaid. Geographic Coverage. Nine of the study states operate statewide Medicaid managed care programs, while six operate capitated arrangements only in designated counties. Capitated systems are often introduced first in states' larger, urban counties and their surrounding areas. Benefits. All of the study states include basic required Medicaid benefits, including physician services, inpatient and outpatient hospital care, primary and preventive care, and the Early and Periodic Screening, Diagnosis, and Treatment program (EPSDT) in their capitated systems. (Two states, Connecticut and California, also offer partially capitated plans, which include all but inpatient services in their capitation rates, as well as fully capitated plans. Several other states, including Oregon and Delaware, had offered this option at one time but have stopped.) Services that may be excluded from plans' benefit packages include specialty services for children enrolled in the state's Title V Children with Special Health Care Needs program; behavioral health services; Part H Early Intervention services; and organ transplants. (Table 2 presents the services that are excluded from the capitation rates paid to managed care plans; these services may be reimbursed on a fee-for-service basis or financed through other programs.) Four states give plans the option of providing certain categories of service, such as prescription drugs or dental services; if plans choose to provide these services, their capitation rates would be enhanced accordingly. B. Rate Development The general methods used by the study states to contract with plans and establish capitation rates can be classified according to two criteria: how the state uses technical criteria to select plans for contracts, and how capitation rates are developed. Within each of these areas, the study states may again be grouped into two categories. For both its technical and price criteria, a state may either develop a set of standards that all contracting plans must meet, or may score plans based on their technical and cost proposals. Health Systems Research, Inc. Chapter II Page 5 Table 2. Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package* Year Program Began Arizona: Mandatory 1115(a) Statewide -AFDC and related None Behavioral health services for all Arizona Health -SOBRA pregnant women & but children age 18-20 who are not Care Cost infants (< age 1) < 140% FPL eligible for the Seriously Mentally Containment III program, non-emergency adult -children age 1-6 < 133% FPL System dental care, specialty services for children covered under the (AHCCCS) -children < age 14 < 100% FPL Children's Rehabilitative Services -food stamp children < age 14 Program, nursing facility services 1982 -SSI, SSI-related, and MN/MI over 90 days, and translation services. California: Mandatory 1915(b) 12 counties -AFDC and related None All behavioral health services, Two-Plan specialty services for children with -pregnant women & infants < Model 200% of FPL special health care needs, major organ transplants, ophthalmic -children age 1-6 < 133% of FPL lenses, and long-term care services -children under age 14 < 100% > 2-month limit. of FPL -SSI, SSI-related, and MN/MI (excluding spend-downs) 1996 Connecticut: Mandatory 1915(b) Statewide -AFDC and related None School-based child health services for CSHCN; inpatient services Connecticut -SOBRA pregnant women & Access infants < 185% of FPL excluded from partially capitated plans -children < age 19 < 185% of FPL 1995 Health Systems Research, Inc. Chapter II Page 6 * Reflects whether enrollment in capitated plans is mandatory or voluntary. ** All study states include required Medicaid benefits such as primary and specialty physician services, outpatient hospital care, EPSDT services, emergency transportation, and, in fully capitated plans, inpatient hospital services in their capitated benefit packages. Table 2. (cont.) Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package* Year Program Began Delaware: Mandatory 1115(a) Statewide -AFDC and related Uninsured adults Pharmacy services, all dental with incomes < Diamond State -SOBRA pregnant women & services, non-emergency 100% of FPL Health Plan infants < 185% of FPL transportation, behavioral health services above limits, private duty -children age 1-5 < 133% of FPL nursing > 28-hour-per-week limit, -children < age 19 < 100% of & skilled nursing facility services FPL > 20-day limit. -SSI eligibles 1996 Florida Voluntary 1915(b) Statewide -AFDC and related None Child & adult dental benefits and -SOBRA children transportation are optional for plans -Foster Children -SSI Medicaid Only -SSI Medicare Part B Only 1982 -SSI Medicare Parts A&B Health Systems Research, Inc. Chapter II Page 7 Table 2. (cont.) Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package** Year Program Began Georgia: Voluntary N/A 24 counties -AFDC and related None All dental services, all behavioral Georgia Better -infants < 185% of FPL health services, non-emergency Health Care transportation, long-term care, and Program -children age 1-6 < 133% of FPL chiropractic services. -children < age 19 < 100% of FPL 1996 -SSI eligibles Massachusetts: Voluntary 1915(b) Statewide -AFDC and related Will expand to All dental services, chronic-care MassHealth households <400% -SOBRA pregnant women & hospitalization, non-emergency of FPL who infants < 185% of FPL transportation skilled nursing receive unemploy- facilities > 100-day limit, and -children age 1-6 < 133% of FPL ment benefits and durable medical equipment > -children under age 14 < 100% of to unemployed $1,500-limit. FPL persons < 133% of FPL. -SSI, SSI-related, and MN/MI (not in an LTC institution) Pharmacy services are optional for plans. 1992 Health Systems Research, Inc. Chapter II Page 8 Table 2. (cont.) Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package Year Program Began Minnesota: Mandatory 1115(a) 16 counties -AFDC and related Infants and Transportation services for those Prepaid children < age 2 -SOBRA pregnant women & enrolled in non-metro area plans Managed Care <275% of FPL infants < 185% of FPL are paid on a fee-for-service basis. Program (PMAP) -children age 2-6 < 133% of FPL -children under age 14 < 100% of FPL -SSI elderly 1985 -medically needy children Missouri: Mandatory 1915(b) 4 regions -AFDC and related None Mental health services for Foster Managed Care children, mental health services > Plus -SOBRA pregnant women & infants < 185% of FPL limit for all other eligibles, and organ transplants. -children age 1-6 < 133% of FPL -children < age 19 < <100% of FPL -Foster Care children (for 1995 physical health services only) Health Systems Research, Inc. Chapter II Page 9 Table 2. (cont.) Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package** Year Program Began New York: Voluntary in 1915(b) 32 counties -AFDC and related None Outpatient mental health services all but two in two Partnership -SOBRA women & infants > 20-visit limit, outpatient counties counties Plan <185% of FPL substance abuse services > 60-visit limit, and inpatient mental health -children age 1-6 <133% of FPL and substance abuse services > 30- -children age 7-16 < 100% of day combined total. FPL -MN/MI (excluding spend- Family planning, dental services, downs) and non-emergency transportation -Home Relief eligibles services are optional for plans. 1996 Oregon: Mandatory 1115(a) Statewide -AFDC and related uninsured adults, Services below line 578 on the Oregon Health -SOBRA women, infants & couples, and prioritized list not covered families < 100% of Plan children < age 6 < 133% of FPL FPL -children < age 19 <100% of FPL -foster children 1993 -SSI eligibles Health Systems Research, Inc. Chapter II Page 10 Table 2. (cont.) Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package* Year Program Began Rhode Island: Mandatory 1115(a) Statewide -AFDC and related uninsured pregnant Adult dental benefits, 90% of women and RIte Care -SOBRA pregnant women, mental health > 30-day/ visit limit, infants & children < age 6 children < age 8 mental health for SED&SPMI, <250% of FPL <185% of FPL early intervention over $3000, (copays) organ transplants, nursing facility -children under age 14 < 100% of care over 30 days, non-emergency FPL services provided in the -Medical Assistance Only emergency room. Families < 70%of FPL 1993 Tennessee: Mandatory 1115(a) Statewide -AFDC and related Uninsurable Adult dental, enhanced support individuals TennCare - SOBRA pregnant women & services for high-risk pregnant (TCHIP), and infants < 185% of FPL women, infants, and children, uninsured CSHCN in foster care, outpatient -children age 1-6 < 133% of FPL individuals mental health > 45-visit limit, -children under age 14 < 100% of outpatient substance abuse FPL services > 2-program-per- lifetime limit, mental health visits for -SSI eligibles SPMI -Medically Needy -children in state custody 1994 Health Systems Research, Inc. Chapter II Page 11 Table 2. (cont.) Medicaid Capitated Managed Care Program Characteristics in Study States, as of 1/1/97 State, Program Mandatory Type of Geographic Eligibility Categories Expansion Groups Benefits Excluded from Capitated Name, and or Voluntary* Waiver limits Benefit Package** Year Program Began Texas: Mandatory 1915(b) 35 counties -AFDC and related children < age 18 Prescription drug coverage, State of Texas -SOBRA pregnant women & ineligible for EPSDT dental care, and mental AFDC and related Access Reform health rehabilitation services infants < 185% of FPL (STAR) b/c of applied -children age 1-6 < 133% of FPL income of -children under age 14 < 100% of stepparents & FPL grandparents, infants < 185% of FPL 1993 Washington: Mandatory 1915(b) Statewide -AFDC and related None Dental servics; mental health and Healthy substance abuse treatment Options -SOBRA pregnant women < 185% of FPL separately capitated -children < age 19 < 200% of 1993 FPL Health Systems Research, Inc. Chapter II Page 12 These two general distinctions can be combined to produce four types of contracting methodologies. This classification is displayed in Table 3 and described below. Table 3. Classification of Contracting Methodologies Technical Standards Plans Meet State Thresholds Plans Scored Capitation Rates Plans Accept State's Connecticut California Rates Florida Delaware Georgia Minnesota Oregon Tennessee Plans Present Cost Massachusetts Arizona Proposals Rhode Island Missouri New York Texas Washington No Price Competition. Six states-Connecticut, Florida, Georgia, Minnesota, Oregon, and Tennessee-award contracts to all plans that meet the state's technical standards and are willing to accept the capitation rates determined by the state. Non-price Competition. In this model, the state determines the number of plans needed in each geographic area and plans compete on technical standards to win these contracts. Two study states, California and Delaware, select plans based on competition only on technical criteria; the plan or plans selected must accept the rates developed by the state. Plans Meet Technical Standards and Negotiate Rates. Two states, Massachusetts and Rhode Island, require plans to meet specific technical standards but do not score the plans on their technical proposals. Rather, the submission of an acceptable technical proposal is a prerequisite for the evaluation of the plan's cost proposal, which is used to negotiate a capitation rate. In these systems, there is no predetermined limit on the number of plans Health Systems Research, Inc. Chapter II Page 13 with which the state will contract; thus, the plans are not in direct competition with each other. Plans Scored on Technical and Cost Proposals. The remaining five states in the study-Arizona, Missouri, New York, Texas, and Washington-use systems in which plans are selected based on the level of their proposed capitation rates as well as technical criteria. In these systems, the weight placed on the plans' cost proposals ranges from approximately 8 percent (50 of 600 total points) in Texas to 60 percent in Washington State. In these states, the number of contracts to be awarded is limited; thus, the plans are in direct competition with each other on both technical and cost criteria. State officials reported a number of reasons for choosing one or another of these contracting methods. Under the first two of these models, states develop capitation rates directly, without requiring cost proposals or negotiation with plans. This method is seen as efficient, especially in states such as Oregon and Tennessee, which were required to implement their managed care systems soon after receiving 1115(a) waivers. In addition, this is seen as an attractive option for states that are beginning to implement managed care systems, as state Medicaid agencies are accustomed to setting fee-for-service rates. Finally, state officials point out that, even in competitive bidding systems, the state would need to develop a benchmark against which to compare plans' bids; therefore, a non-competitive system in which the state develops capitation rates seems like a logical first step, even if states plan to implement competitive bidding systems in the future. The major advantage reported for price competition is the potential it offers for cost savings. The desire to win a contract in a competitive bidding system presents a clear incentive for plans to present low bids. In addition, four states have designed algorithms for assigning those eligibles who do not voluntarily choose a managed care plan to favor those plans that offer the lowest capitation rates or receive the highest total point scores. (These systems will be discussed in detail in Section G below.) In addition, in many states, competition itself is valued by state officials and legislators. However, a disadvantage of a competitive process is the potential for significant disruptions to occur in later years if a participating plan or plans lose their contracts and their members are forced to switch to other plans. Health Systems Research, Inc. Chapter II Page 14 The following sections of this report discuss in detail the study's findings in the following areas: The technical standards and criteria used to select plans; The specific methods used to develop rates or UPLs; The systems used to negotiate prices with plans and the results of these processes; The standards used to certify and select plans in states that do not use price competition; Methods used to share risk with plans; Enrollment provisions related to capitation rates; and Quality assurance and monitoring efforts related to capitation rates. C. Technical Criteria Each study state establishes technical criteria for managed care plans contracting with the state. These criteria may take the form of thresholds which all contracting plans must meet, or they may describe areas in which offering plans are scored. Examples of the types of technical criteria that are used in the study states and their role in the contracting process are presented below. In eight study states, technical standards are set that all plans must meet in order to receive contracts. These include standards related to the breadth of provider networks, the accessibility of care, administrative capacity and ability to meet reporting requirements, quality monitoring procedures, availability of emergency care, grievance procedures, and financial solvency. Technical criteria may be the only criteria on which plans compete with each other. For example, under California's most recent managed care initiative, the Two-Plan Model, the state will contract with two managed care plans in each of twelve pilot counties. One of these plans, the Local Initiative, will be a network of existing public-sector and safety-net providers, including county hospitals, local health clinics, and community health centers. The other will be a commercial plan. Thus, the state must select one commercial plan in each pilot county. These plans are chosen based on technical criteria only, including financial solvency, capacity, and overall quality. Plans may gain additional Health Systems Research, Inc. Chapter II Page 15 points by including safety net providers in their networks or by offering additional benefits beyond the required package. Scores on the cost proposal may be combined with scores on technical criteria to develop a total score for each bidder. Examples of the types of technical criteria scored in the study states include the quality, size, and comprehensiveness of the plan's provider network; ability to meet outcome goals, such as improvements in infant mortality and low birth weight rates; commitment to internal quality improvement and ability to provide encounter data; the quality of program administration, including executive management and grievance procedures; the ability of the plan to perform administrative functions, including data collection and reporting; financial stability; cultural and linguistic requirements; and the ability to provide health education services. D. Actuarial Methods As mentioned above, states' methods for establishing capitation rates can be divided into two broad categories: those that establish specific rates to be paid to all contracting plans, or those that require plans to present cost proposals, with final rates agreed upon through negotiation between plans and the state Medicaid agency. In both of these types of systems, state Medicaid agencies are required to develop Upper Payment Limits (UPLs) to guide the selection of and negotiation with plans and to assure that the rates established meet the fiscal guidelines required by their Medicaid waivers. This section presents the study's findings on the following aspects of capitation rate development: Federal guidelines governing the rates paid under each type of Medicaid waiver; The age, sex, and geographic categories used to develop capitation rates and Upper Payment Limits (UPLs); Actuarial methods used to develop capitation rates and UPLs; Adjustments used to both increase and decrease base capitation rates and UPLs; and Specific issues related to the development of rates for pregnant women and newborns. In developing these rates or rate guidelines, the states are restricted by the requirements of 1915(b) and 1115(a) waivers regarding expenditures under these waiver programs. Health Systems Research, Inc. Chapter II Page 16 Demonstrations under 1115(a) waiver authority must be budget-neutral to the federal government over the life of the demonstration. There is no limit on state funding for the demonstration, nor is there any explicit limit on capitation rates paid under these programs; however, any increased cost over the projected federal and state expenditures in the absence of the waiver must be borne entirely by the state. Programs operating under 1915(b) waivers must show cost savings compared to expenditures for similar services provided to an actuarially-equivalent population under the fee-for-service program. Thus, combined federal and state Medicaid expenditures must be reduced under this waiver, and capitation rate must be less than a fee-for-service equivalent rate. Rates and rate guidelines used by states are set using a variety of parameters, including enrollees' age and sex, eligibility category, and geographic location. Table 4 shows the factors used in each study state and the total number of rate cells used for developing rates for AFDC and related populations in the most recent contracting process. The number of cells presented in this table reflects the total number of rates developed for this population; in states where plans are asked to present cost proposals, this represents the number of distinct prices the plans are asked to propose, while in states that determine rates, this is the number of different rates paid. As Table 4 shows, nearly all study states used age and sex groups to develop age- and sex- specific rates or to adjust final rates for the demographic distribution of each plan or geographic area. Twelve states set distinct rates for specific geographic areas within the state, either counties or multi-county regions, to reflect differences in the cost of providing care in different areas of the state. Finally, nine states set distinct rates for specific eligibility groups. Health Systems Research, Inc. Chapter II Page 17 Table 4. Rate Cells State Number of Cells Age/Sex Groups Geographic Areas Eligibility Categories (AFDC and related only) Arizona 45 - age 0-1 M&F 9 areas (1-2 counties in each) -AFDC and related (including - age 1-13 M&F SOBRA women and OBRA - age 14-44 F children) - age 14-44 M - age 45+ M&F California 12 N/A (rates adjusted for each 12 counties -Family (including AFDC and county's demographics) related) Connecticut 48 - age 0-1 M&F 8 counties None - age 1-14 M&F - age 15-39 F - age 15-39 M - age 40+ F - age 40+ M Delaware 8 - age 0-1 M&F statewide -AFDC and related (including - age 1-10 M&F OBRA children) - age 11-17 F -SOBRA women (all ages) - age 11-17 M - age 18-44 F - age 18-44 M - age 45+ M&F Florida 60 - age 0-1 10 areas None - age 1-5 - age 6-13 - age 14-20 - age 21-54 - age 55+ Health Systems Research, Inc. Chapter II Page 18 Table 4. Rate Cells State Number of Cells Age/Sex Groups Geographic Areas Eligibility Categories (AFDC and related only) Georgia 24 AFDC and related: 2 areas -AFDC and related - age 1-2 mos M&F -Right from the Start Medicaid - age 3-12 mos M&F (OBRA) children - age 1-6 M&F - age 7-13 M&F - age 14-44 M - age 14-44 F - age 45+ M&F OBRA children: - age 0-1 M&F - 1-6 M&F - 7-13 M&F - 14+ M - 14+ F Massachusetts 3 N/A (rates adjusted for each 3 areas -AFDC and related plan's demographics) Minnesota 24 - age 0-1 M 3 areas None - age 0-1 F - age 2-15 M - age 2-15 F - age 16-49 M - age 16-49 F - age 50+ M - age 50+ F Health Systems Research, Inc. Chapter II Page 19 Table 4. Rate Cells State Number of Cells Age/Sex Groups Geographic Areas Eligibility Categories (AFDC and related only) Missouri 36 - age 0-1 M&F 4 areas -AFDC and related - age 1-6 M&F -SOBRA pregnant women - age 7-13 M&F - age 14-20 M - age 14-20 F - age 21-44 M - age 21-44 F - age 45+ M&F New York 45 - age 0-6 mos M&F 9 areas -AFDC and related - age 6 mos-20 yrs M - age 6 mos-14 yrs F - age 15-20 F - age 21-64 M&F Oregon 15 N/A 5 regions -OHP Basic (AFDC and related Families<100%FPL, Adults and Couples<100%FPL, Poverty Level Medical Adults and Children <100%FPL) -PLM Adults<133%FPL, -PLM Children<133%FPL Rhode Island 6 - age 0-1 M&F statewide None - age 1-5 M&F - age 6-14 M&F - age15-44 M - age 15-44 F - age 45+ M&F Health Systems Research, Inc. Chapter II Page 20 Table 4. Rate Cells State Number of Cells Age/Sex Groups Geographic Areas Eligibility Categories (AFDC and related only) Tennessee 6 - age 0-1 M&F statewide None - age 1-13 M&F - age 14-44 M - age 14-44 F - age 45-64 M&F - age 65+ M&F Texas 36 N/A 6 areas (5-7 counties each) -AFDC adults -AFDC children -Pregnant women -Newborns -Children under age 19 <100% FPL -Medicaid-only (AFDC non- cash) children Washington 72 - age 0-1 M&F 8 regions None - age 1-5 M&F - age 6-18 M - age 6-18 F - age 19-34 M - age 19-34 F - age 35-64 M - age 35-64 M - age 65+ M&F Health Systems Research, Inc. Chapter II Page 21 In five of these, all AFDC and related categories, including pregnant women and children in expansion categories, are grouped in one eligibility category for rate development purposes. In the remaining four, separate rates may be set for pregnant women in the SOBRA eligibility group or children in the OBRA expansion category, or separate rates may be developed for each individual eligibility category. Overall, four states used demographics, geography, and eligibility category as parameters for the development of rates, while nine states used two of these three criteria. Thus, the total number of rate cells used ranged from three in Massachusetts (where one Upper Payment Limit was developed for each of the state's three regions within the AFDC and related category) to 72 in Washington (where rate guidelines were determined for nine age/sex groups in eight geographic regions). In general, the advantage of consolidating rates or Upper Payment Limits is the simplicity of their calculation; smaller divisions of the eligible population produce more specific, and thus potentially more accurate, rates, but are administratively more complex. In most states, the rates within each cell are calculated using databases of past years' fee-for- service claims. From these raw databases, states generally first subtract claims for population or services excluded from the Medicaid managed care program; additional Medicaid expenditures, such as Disproportionate Share Hospital or Graduate Medical Education payments, are generally subtracted at this point as well. The total remaining expenditures within each category are then divided by the number of enrollee months represented by the category to develop base per-member-per-month (PMPM) expenditures. In every study state, these base calculations are subject to a number of adjustments, including both increases and decreases in the rates, in the process of arriving at final rates or UPLs. In nine study states, these analyses were conducted by outside actuaries, although officials of state Medicaid agencies were frequently closely involved in the development of rates and rate guidelines. Examples of adjustments used by the study states are presented below. 1. Increases Base PMPM rates may be increased by a variety of factors. Common factors used in the study states include the following: Health Systems Research, Inc. Chapter II Page 22 Trending to Contract Year. Because the base rates are generally based on past years' fee-for-service data, all study states apply inflation or trend factors to their base PMPM rates to trend these rates forward to the contract year. These trend factors generally account for both increases in Medicaid reimbursement rates and observed changes in service use. Nine states used a single overall trend factor, while six developed service- specific factors to account for expected or observed changes in utilization or reimbursement rates. In some states, such as California and Georgia, these factors were intended to account for all service-specific changes in utilization, including expected decreases; thus, these factors may result in reductions in rates for specific service categories. In other states, such as Delaware, these factors account specifically for inflation, and thus always result in rate increases. Administrative Costs. Nine states included a factor to account for the cost to the plan of the administrative aspects of providing care to the Medicaid population. These factors ranged from 2 percent in California to 10 percent in Georgia. In some states, such as Massachusetts, these factors were intended to approximate the percentage of Medicaid expenditures devoted to administration; thus, these states used administrative cost factors in the range of 2 to 3 percent. In other states, such as Oregon and Georgia, these factors were intended to approach the administrative cost factor generally found in managed care plans; thus, these states used higher rates. Six states did not include an explicit factor to compensate plans for the cost of administering the Medicaid managed care program. Completion Factors. The fee-for-service claims databases on which rates and UPLs are generally based are frequently incomplete; that is, services may have been provided for which Medicaid has not received invoices, and claims may have been received that have not been paid, at the time the actuarial database is created. Therefore, six states apply completion factors to increase calculated rates by the amount estimated to be missing from the claims database; these factors generally amount to less than 2 percent. Adjustment for Pent-up Demand. One state, Washington, conducted a study of Medicaid claims experience that showed that new enrollees joining the state's Medicaid managed care program often postpone receiving needed care until they have been enrolled in a managed care plan, although services are available on a fee-for-service basis from the time of enrollment. Thus, an increase was applied to the capitation rates to compensate plans for higher-than-expected utilization during these months. Health Systems Research, Inc. Chapter II Page 23 2. Decreases Many states also apply a number of factors that reduce the base PMPM rates. These include adjustments to preserve savings and rebates the states had received under the fee-for-service system as well as adjustments to account for savings anticipated under managed care. These factors may include the following: Prescription Drug Rebate Adjustment. Four states reduce their capitation rates to account for the savings that had been gained under the Federal Drug Rebate Program. Under the fee-for-service system, this program allowed state Medicaid agencies to received discounted prices on prescription drugs in the form of rebates; the expenditures in the claims databases, however, represent the cost to the state before the rebate. To assure that the state will pay the same net amount for pharmacy services under managed care, a discount, generally amounting to approximately 18 percent, is taken from these gross pharmacy expenditures. Adjustment for Third-party Liability. Since Medicaid is the payer of last resort for its enrollees, providers are required to bill any other insurers who cover Medicaid eligibles before billing Medicaid. In most study states, fee-for-service claims databases do not include payments received from third-party insurers. However, seven states include a deduction to represent expected receipts from third-party insurers, either in their UPL calculations or in the guidance given to plans regarding the calculation of their bids. In Missouri, plans may opt to collect from third-party insurers, although they are not required to do so. If they do not (and most do not), their rates will be increased slightly to compensate for these funds, and collections will be made by the state Medicaid agency. Adjustment for Voluntary Selection. In four study states, enrollment in the capitated managed care program is optional; Medicaid eligibles may choose a fee-for-service PCCM model if they prefer. Three of these states include a factor in their capitation rates to account for the assumption that those who choose the HMO option are likely to be slightly healthier and less likely to use services than the Medicaid-eligible population as a whole. Adjustment for Uncompensated Care. The state of Tennessee includes in its rate calculations an adjustment to preserve the contribution previously provided by the state's hospitals through charity care. Since TennCare provides coverage for a substantial number of previously uninsured patients, the state reduced the rates paid to plans by 22 percent to assure that hospitals did not receive windfalls once these patients obtained insurance. Health Systems Research, Inc. Chapter II Page 24 Adjustment for Lost Interest. The state of California includes in its rate calculations a factor to account for interest lost to the state through prepayment of capitation rates at the start of each month. Under fee-for-service reimbursement, the state Medicaid agency would receive invoices throughout each month; interest would accrue on Medicaid funds for those invoices received or paid at the end of the month. The state's rate calculation methodology therefore includes a reduction (of less than 1 percent) to compensate the state for this lost interest. Managed Care Savings. Finally, many states developed their capitated managed care initiatives with the intention of reducing total Medicaid expenditures or generating sufficient savings to significantly expand eligibility for the program. Thus, seven states include in their calculations an explicit factor representing the savings the state hopes to achieve through the program. These factors range from a 3 percent reduction in Washington to 5 percent in Minnesota and Connecticut. (In several states, such as Texas, the amount of the managed care discount is not public information.) Instead of an across-the-board discount, four states applied service-specific estimates of the effect of managed care on service use and expenditures to the assumptions included in the actuarial models. Table 5 on the next page shows the assumptions used by these states. As the table shows, assumptions may be made regarding the effect of managed care on utilization of specific services, on the unit costs of services, or on total expenditures in specific service categories. The adjustments to utilization rates are based on the assumption that managed care will be successful in managing the use of high-cost services and in substituting lower-cost alternatives for inpatient and emergency room care. The adjustments to unit cost projections are based on managed care plans' ability to negotiate discounts with providers such as physicians and pharmacies. The adjustments to overall expenditure attempt to take both of these phenomena into account using a single adjustment figure. Health Systems Research, Inc. Chapter II Page 25 Table 5. Service-Specific Adjustments to Account for Anticipated Effects of Managed Care State Adjustment Inpatient Outpatient Emergency Primary Other Pharmacy Mental Health Maternity "Other" Room Care Physician Services Physician Delaware Utilization - 18% to -4% to -20% +10% N/A N/A - 5% (< age 18) N/A 0% - 25% - 35% to - 38% (> age 18) Georgia Unit cost - 10% 0% 0% 0% - 5% - 5% N/A N/A - 5% Utilization 15% + 5% - 20% + 5% - 10% - 10% N/A N/A - 5% Missouri* Expenditures - + - + N/A N/A N/A N/A N/A New York* Expenditures - + - + N/A N/A N/A N/A N/A Oregon Expenditures 30% - 30% N/A + 10% + 10% + 5% N/A - 10% N/A *The magnitude of adjustments used is not public information. Health Systems Research, Inc. Chapter II Page 26 3. Demographic Adjustments In several states, rates paid to plans in each geographic area are adjusted for the demographic makeup of Medicaid enrollees in that area. This type of adjustment is generally made in states that do not set separate rates for specific age and sex categories. These adjustments take into account the distribution of enrollees in each region across age and sex categories. In some cases, similar adjustments are made to account for the distribution of enrollees across specific eligibility categories within each of the larger eligibility groupings on which rates are based. These adjustments, often described as "relational modeling," involve reducing the rates paid to plans operating in areas with lower-cost populations and increasing those paid in costlier regions. Thus, although the rates paid in individual cells will be increased or decreased, the net change in the total capitation amount paid is zero. 4. Special Provisions for Delivery and Newborn Costs In developing capitation rates and rate guidelines, many states have encountered difficulties in setting rates for pregnant women and newborns. In many fee-for-service systems, for example, services provided to newborns are billed on the mother's Medicaid number, as the infant cannot be enrolled independently immediately upon delivery. In order to accurately estimate expenditures for newborns and set rates within this age group, two states, Georgia and Washington, shift these charges from the mothers' rate category (based on age, sex, and eligibility group) to that of infants. The issue of setting rates for pregnant women, particularly those in the SOBRA eligibility category, is more complex. These women do not become eligible for Medicaid until they become pregnant, and often do not enroll in Medicaid until they begin prenatal care, which may be relatively late in pregnancy. Moreover, their eligibility ends 60 days after the month in which they deliver. Therefore, plans may incur large expenses for delivery and receive relatively few months' worth of capitation payments with which to finance these expenses. States have addressed this problem in a variety of ways, as described below. Two states in which enrollment in capitated systems is voluntary, Georgia and Florida, exclude SOBRA women from enrollment in capitated systems. Health Systems Research, Inc. Chapter II Page 27 Six states (Arizona, Delaware, Missouri, New York, Rhode Island, and Washington) pay plans on a per-case basis for deliveries, in addition to monthly capitation payments. In these states, separate capitation rates are not calculated for SOBRA women; rather, plans receive the rate for AFDC and related women of childbearing age on behalf of these women. In the states in which plans present cost proposals (Arizona, Missouri, New York, Rhode Island, and Washington), the plans are asked to develop the rate to be paid per delivery, generally based on an average of the cost of vaginal and cesarean deliveries, weighted according to the expected proportion of deliveries attributable to each method. This base amount is then adjusted to account for the fact that the base capitation rate paid for these women's care includes some likelihood of pregnancy and delivery, as some proportion of AFDC-eligible women give birth each year. Therefore, some of the cost of maternity-related services are included in the base capitation rate. To avoid double-payment, the proposed capitation rate for five or six months is subtracted from the proposed delivery payment amount. This supplemental payment will either be paid at the time of delivery, as in Arizona, or at the time of enrollment of a SOBRA pregnant woman, as in Delaware. Arizona is unique in providing this payment only for deliveries to SOBRA women; the other states provide this payment for all deliveries to Medicaid-eligible women. Missouri officials explained that while the payment was originally intended to apply only to women in the SOBRA category, the determination of eligibility category for each delivery has become too administratively complex; therefore, the state is planning to begin making the supplemental payment to plans for all Medicaid deliveries in the next contract year. Oregon has developed a slightly different system of reimbursing plans for delivery costs. The state has established a "maternity and newborn care withhold pool," in which 25 percent of the capitation amount attributed to maternity and newborn services for the relevant categories of pregnant women and children is deposited. The funds in this pool are then divided among participating plans every six months, based on the distribution of deliveries across plans. Thus, in this system, no additional funds are provided to plans for delivery services; funding is simply redistributed among plans based on the number of deliveries provided. E. Negotiation Methods The previous section described processes used by the study states to develop either a set of rates to be paid to contracting plans or a set of guidelines against which to compare plans' cost proposals. This section focuses on the methods used to negotiate rates in those study states that Health Systems Research, Inc. Chapter II Page 28 require plans to submit cost proposals. Seven of the 15 study states require plans to present cost proposals, which are compared to the state's rate guidelines. These guidelines always include an upper limit, often based on the Upper Payment Limit required by 1915(b) waivers, and may also include explicit lower limits, which represent the lowest rate state officials feel would be adequate to support the provision of comprehensive, high-quality care. Both the upper and lower payment limits may or may not be shared with plans through the Request for Proposals. The limits developed and shared with plans in the seven relevant states are presented in Table 6. The sections below describe, in turn, these seven states' systems for soliciting bids and comparing them to rate guidelines, and for scoring cost and technical proposals and selecting plans to receive contracts. Table 6. Bid Solicitation Processes in States Requiring Cost Proposals State Category Upper Limit Lower Limit Arizona Plans present cost proposals Explicit, not shared Explicit, not shared Massachusetts Plans present cost proposals Explicit, not shared Implicit, not shared Missouri Plans present cost proposals Explicit, not shared Explicit, not shared New York Plans present cost proposals Explicit, not shared Explicit, not shared Rhode Island Plans present cost proposals Explicit, shared Explicit, shared Texas Plans present cost proposals Explicit, shared Implicit, not shared Washington Plans present cost proposals Explicit, shared Explicit, shared 1. Bid Solicitation Processes The first step in the contracting process, after the development of upper (and, in some cases, lower) payment limits, is the development of a Request for Proposals to be distributed to potential offerors, outlining the state's technical requirements and the process for developing proposed capitation rates. As discussed above, this RFP may or may not include the state's rate guidelines. This section discusses the types of rate guidelines set by the study states and their policies for sharing these guidelines with offering plans. Health Systems Research, Inc. Chapter II Page 29 a. Explicit Upper and Lower Limits Shared with Plans Rhode Island and Washington set explicit upper and lower limits and share them with plans. These limits are determined based on the rate guidelines calculated by the state or its actuaries, as described below: Rhode Island, which operates its RIte Care program under an 1115(a) waiver, is not bound by federally-enforced upper payment limits. Therefore, the state calculates its adjusted fee-for-service equivalent expenditure and sets this as the midpoint of a rate range. The upper and lower limits of the range are set at 7.5 percent above and below this midpoint. The State of Washington sets a target rate equal to an adjusted fee-for- service equivalent expenditure and a lower limit at 90 percent of the target. The advantage of sharing rate guidelines with plans is in both the atmosphere of openness it creates in the relationship between the state and the plans, and in the efficiency that it creates in the negotiation process. If rate guidelines are known to the plans, the likelihood is great that the plans will respond with rates that fall within the acceptable range, thus limiting the need for counter-offers or best and final offer rounds. The clear disadvantage of sharing upper and lower limits with plans is that the state runs the risk of receiving only bids at the top of the range. States may mitigate this problem by developing a scoring system that rewards plans for offering low bids, as Washington has done. Even in this system, however, the final rates fell near the top of the range, clustering around 97 to 99 percent of the target rate. Rhode Island was not able to take advantage of the opportunity to establish incentives. In a small state with few licensed HMOs, state officials felt they had to accommodate plans' stated needs; thus, after the first round of bidding, the state was required to raise the upper and lower limits of their range to meet plans' bids. Even after this adjustment, the final contract rates fell at the top of the rate range. Health Systems Research, Inc. Chapter II Page 30 b. Explicit Upper and Lower Limits Not Shared with Plans To avoid offering plans an incentive to bid at the top of a range, three states-Arizona, Missouri, and New York-set explicit upper and lower payment limits but do not share them with plans. (The breadth of these ranges or the magnitude of adjustments made in developing them are not public information in these states.) In these systems, the plans' cost proposals are compared to the state's acceptable range of rates and given a score; if plans bid above or below the range, the state develops a counter-offer based on a standard algorithm, and scores the plan based on the counter-offer. The negotiation systems used in these three states are described below. Arizona solicits bids from plans, after an initial evaluation, they are allowed to adjust any rates that fall above or below the state's ranges during a Best and Final Offer round. In this round, the plan will be told which specific rates should be increased or decreased, although the total amount of the bid may not be increased; if the rate in any cell is raised, another must be correspondingly reduced. After the last bid is offered, the state will make counter-offers in any cell for which the offerors' rate still does not meet the state's limits. For all rates below the lower limit, the plan will be offered a rate equal to the lower limit, and for all rates above the limit, the counter-offer will fall in the bottom half of the range. This final provision serves as an incentive for plans to try not to exceed the range with their bids. Missouri receives plans' bids and scores them according to their relationship to the state's upper and lower limits and their place in the range of bids received. If the state chooses, it may conduct a Best and Final Offer (BFO) round in which plans are told which rates fall above or below the allowable range. However, as soon as the state successfully negotiates contracts with a sufficient number of plans in a given region, no further negotiations will be conducted with other plans; therefore, plans are encouraged to approach each bid as its best and final offer. New York evaluates the rates of only those plans who are deemed "qualified" to serve Medicaid eligibles based on their technical proposals. The state's process for evaluating cost proposals and negotiating with plans is similar to Missouri's, with the state requesting Best and Final Offers when it is in its interest to do so, and accepting no further offers after a sufficient number of plans have been engaged. If, after the BFO round, rates in certain cells still fall outside the range, the state will develop a counter-offer. For rates that fall below the lower limit of the range, a rate equal to the lower limit will be offered; for rates Health Systems Research, Inc. Chapter II Page 31 that fall above the range, the counter-offer will fall below the top quartile of the range. Plans that agree to these counter-offers are then eligible to receive contracts. However, the actual awarding of contracts is based on county officials' determination of the number of plans to be engaged in the county. Once this determination has been made, the plans are ranked according to their scores on the technical proposal (not on their negotiated rates), and the plans with the highest scores in each county will receive contracts. Because little information about the results of these bidding processes is public, officials of these states were unable to discuss their states' success in soliciting bids substantially below the upper limits of the states' rate guidelines. C. Explicit Upper Limit Shared with Plans Texas sets an upper payment limit and shares it with plans as part of the Request for Proposals. This limit is described as providing a degree of cost savings that is satisfactory to state officials; therefore, although sharing the upper limit with offering plans creates the incentive to produce bids at the limit, this is not seen as a disadvantage. In fact, although the level of plans' bids is scored as part of the bidding process, the cost proposal only represents approximately 8 percent of the total available points. d. Explicit Upper Limit Not Shared with Plans Massachusetts is the only study state that sets only an upper payment limit that is not shared with plans. The state has historically struggled to contract with plans for rates that fall below this limit, so state officials have not felt it necessary to set an explicit lower bound for capitation rates. 2. Scoring Systems In each of the states described above, cost proposals are scored according to comparisons with the rate guidelines. These scores may be combined with scores on a technical proposal, or the submission of an adequate set of rates may be a prerequisite for the consideration of the Health Systems Research, Inc. Chapter II Page 32 technical proposal. Table 7 shows the proportion of points allotted to the cost proposal in each state; these ranged from approximately 8 percent in Texas to 60 percent in Washington. In three states, the plans were required only to negotiate rates that fell within the state's guidelines. In Massachusetts and Rhode Island, all plans that successfully negotiate rates and meet the state's technical criteria are awarded contracts. In New York, the submission of an acceptable technical proposal is a prerequisite for the evaluation of the cost proposal. The technical standards that plans must meet concern the quality and depth of provider networks; the use of traditional Medicaid providers, including school-based health centers; the accessibility of providers and services; plans' quality assurance mechanisms; data and reporting systems; complaint resolution processes; and timeliness of provider payments. Table 7. Scoring Systems in States with Price Competition State Contracting Method % of points based on price Arizona Plans selected based upon level of proposed 28% capitation rate and other non-price factors. Massachusetts Plans selected based upon level of proposed Rates must fall under upper capitation rate and other non-price factors. limits Missouri Plans selected based upon level of proposed 40% capitation rate and other non-price factors. New York Plans selected based upon level of proposed Rates must fall within ranges capitation rate and other non-price factors. Rhode Island Plans selected based upon level of proposed Rates must fall within ranges capitation rate and other non-price factors. Texas Plans selected based upon level of proposed 8.3% capitation rate and other non-price factors. Washington Plans selected based upon level of proposed 60% capitation rate and other non-price factors. 3. Re-Enrollment Provisions In states in which plans compete with each other for Medicaid managed care contracts, the possibility exists that plans that successfully win contracts in one year will lose these contracts when they are next re-competed. This may cause several problems. First, enrollees' continuity Health Systems Research, Inc. Chapter II Page 33 of care will be disrupted if they are required to leave one plan and join another; they may lose access to providers to which they have become accustomed, and they may not receive needed services before they have become familiar with a new plan's systems. Second, disenrolling and re-enrolling large numbers of people in a short period of time may be an operational challenge for state or county agencies responsible for Medicaid enrollment. If the agency cannot re-enroll all of a plan's enrollees in the month after a contract ends, these enrollees may lose access to care completely. Several states have developed strategies for addressing these issues. These include the following: Washington determined that no more than one third of Healthy Options enrollees in a county be required to disenroll and re-enroll at the end of a contract period. Thus, any plan that enrolled at least one-third of eligibles in a county was essentially guaranteed to win a contract during the next year's competition, as long as its bid was within the acceptable range. (Interestingly, although plans were aware of this provision, not all of the plans that were guaranteed contracts in the most recent contracting period submitted bids at the top of the range.) Arizona has allowed two options when existing plans lose contracts; the plan's enrollees may be disenrolled and required to enroll in a new plan, or the plan's enrollment may be capped, allowing the plan to continue to serve its current enrollees but not to enroll new Medicaid members. If a plan's enrollment is capped, the plan will be paid the final rate bid during the current contracting process (provided that rate fell within the acceptable range). In general, AHCCCS officials choose this option in cases in which the plan is large enough that disenrolling its members would disrupt the continuity of care (but not if continuing their enrollment would threaten the quality of care). Massachusetts and Missouri require plans that lose contracts to continue to serve their members until they can all be re-enrolled in new plans. While these enrollees remain in the plan, the plan continues to receive the previous year's capitation rates. In choosing a mechanism for handling plans that lose contracts, states must balance the competing goals of cost containment and continuity of care for the plans' enrollees. The approach used by Massachusetts and Missouri favors the former, and Washington's the latter; Arizona has chosen to make this decision separately for each individual plan. Health Systems Research, Inc. Chapter II Page 34 F. Risk and Profit Sharing The most common mechanism for sharing risk with plans is through the availability of state- funded reinsurance. Of the study states, five offer state-funded reinsurance to plans; two additional states offered this reinsurance in the past, but have stopped, since plans preferred to buy such insurance on the private market. When plans do accept state-sponsored reinsurance, their capitation rates will be reduced by an amount actuarially equivalent to the value of the benefit. An additional eight states require plans to maintain private reinsurance. Four states reported that they have specific additional mechanisms for sharing risk or profits with plans; in two cases, these arrangements are provided to protect plans organized by traditional safety-net providers. These arrangements are described below: Rhode Island has developed a "risk corridor" arrangement with Neighborhood Health Plan of Rhode Island (NHPRI), the plan made up of the state's community health centers. Under this agreement, the plan bears full risk for expenses amounting to 88 percent of capitation payments; above this limit, the plan is fully responsible for only 1 to 3 percent of expenses incurred, depending on the plan's total enrollment. Responsibility for the remaining 97 to 99 percent of expenses is shared with the state, with the state paying a decreasing portion over time, from 90 percent for the first three quarters of the 1996-97 contract to 60 percent in the 1997-98 contract year. Profit will be shared on an annual basis depending on the net profit achieved by NHPRI, with the amount of profit to be shared with the state to be discussed between the two parties. Under its Two-Plan Model, California has established risk-sharing arrangements with the Local Initiatives (LIs), plans made up of each county's public-sector providers, for the expenditures they make to Federally Qualified Health Centers (FQHCs). Under these arrangements, a risk corridor is established between 90 and 110 percent of the portion of the capitation payment attributed to FQHC costs. If an LI's payments to FQHCs amount to less than 90 percent of payments, the difference is recovered by the state; if expenses exceed 110 percent of the expected expenditure, the state will reimburse the plan for these excess expenditures. Texas requires plans participating in the STAR program to share any profits they derive from the program equally with the state for the first two years of their contracts in each Service Area. This requirement was intended to provide a disincentive for plans to profit from participating in the Medicaid managed care program. Health Systems Research, Inc. Chapter II Page 35 Connecticut, one of the few states that continues to offer partially-capitated plans in its Medicaid managed care system, has developed a profit-sharing system with these plans. Rather than the state requesting a share of the plan's profits, however, this system requires the state to share evenly with the plans any savings realized in the area of inpatient care. Although the partially- capitated plans are not required to finance inpatient services, they are asked to manage the use of these services; thus, this arrangement counters the plans' incentive to admit patients to inpatient care to reduce their own costs. Two states, Tennessee and New York, currently have methods for compensating plans for serving high-risk or high-cost eligibles. (Other states, such as Minnesota, report that they are in the process of developing risk-adjustment methodologies.) In New York, plans are paid an enhanced rate if certain high-risk groups-including those with symptomatic AIDS, asymptomatic HIV-positive enrollees, and those who are diagnosed as Seriously and Persistently Mentally Ill (SPMI)-are disproportionately represented in the plan's enrollee population. For example, if the enrollees with AIDS in any plan account for more than .1 percent of the plan's total Medicaid enrollment, the plan will receive enhanced capitation rates. Tennessee's TennCare program compensates plans for each enrollee who meets a definition of "high risk" developed by the state Medicaid agency, including those with AIDS, organ transplants, kidney or heart disease, and pregnant women. The state sets aside funds each year (to date, these funds have amounted to $40 to $55 million each year), to be divided among plans based on the number of high-risk enrollees in each plan in each quarter. G. Enrollment Provisions Related to Capitation Rates As mentioned above, in states that develop capitation rates through price competition, the enrollment of those eligibles who do not voluntarily choose a managed care plan may be used as an incentive for plans to present low bids. In states that set uniform rates for all participating plans, other methods are used to assign non-choosers to plans. The methods found in the study states to enroll those who do not voluntarily choose a plan are described below: Four states, all of which require participating plans to accept rates set by the state, assign enrollees randomly to a plan in their region or county. These states are Oregon, California, Minnesota, and Tennessee. Health Systems Research, Inc. Chapter II Page 36 Two states in which enrollment in capitated plans is voluntary, Georgia and New York, automatically enroll non-choosers in fee-for-service PCCM programs, so no one is enrolled in an HMO involuntarily. The remaining nine states have developed mechanisms for enrolling those who fail to choose a plan that give preference to particular plans based on specific criteria, as described below. - Price. Rhode Island uses the level of capitation rates to give preference to specific plans. Plans' rates are developed using price competition and negotiation; the resulting rates to develop an algorithm for the assignment of non-choosers to plans. In this algorithm, the lowest bidder receives a plurality of the assigned population; the exact proportion of non-choosers assigned depends on the total number of contracted plans. If four plans participate, for example, they would be ranked in descending order by capitation rate and assigned 40, 30, 20, and 10 percent of assignees, respectively. State officials believe that this policy provides an incentive for plans to present low bids, as those who do not choose a plan are assumed to be relatively healthy; those with high levels of need are more likely to have a relationship with a provider and to select the plan that includes that provider in its network. Quality. Delaware, which develops standard rates but requires plans to compete based on technical quality, gives preference in assigning enrollees to plans with the highest technical score. Relative weights are developed for plans in each area of the state based on the plans' scores on the technical proposal, the degree to which their provider networks include traditional Medicaid providers, and their past success at reaching EPSDT screening goals. These weights then determine the percentage of non-choosers in each area which will be enrolled in each plan. Price and quality. Missouri and Washington used algorithms that favor plans with the highest total evaluation scores, including both the technical and cost proposals. In both states, the final proportions developed for the plans reflect both the rankings of plans by total score and the relative spread between plans' scores. The order in which plans receive preference for enrollment of the assigned population is based on the total points each received, with the highest-scoring plan receiving the largest percentage of assigned enrollees. The actual percentage distribution of enrollees to plans is based on the difference in the number of points each plan is awarded; thus, if plans' scores are close to equal, the assigned population would be distributed approximately evenly, while if one plan far outscored the others, it would receive a substantially larger proportion of the assigned population. Connecticut, which sets uniform rates for participating plans, has developed a separate bidding process to select the Designated Plans to Health Systems Research, Inc. Chapter II Page 37 which non-choosers will be assigned. This competition is open only to plans that have been certified through the general contracting process; bidders are then required to meet an additional set of quality standards and to submit a cost proposal that presents a discount to be taken from the capitation rates paid for all members, not just the assigned population. The technical criteria included in this RFP process are based on the assumption that the assignees are generally healthy but need additional information and outreach to encourage them to use services. The upper limit for the cost proposal is 95 percent of the state's standard capitation rates. Enrollment targets. Arizona sets an enrollment target for each plan, respresenting the percentage of the total access population each is to enroll. These targets are used to guide assignment of non-choosers to plans. That is, the plan whose enrollment is farthest from the stated target will receive preference each time an enrollee is to be assigned to a plan. The targets themselves are based in part on plans' cost and technical proposals; the state assigns each plan a target percentage of the total eligible population, with the highest percentages assigned to plans receiving the highest scores in both areas. Popularity among choosers. In three additional states, Florida, Massachusetts, and Texas, the distribution of plan and program choices among choosers guides the distribution of assignments. In Texas, where enrollment in the capitated STAR program is mandatory, the goal of the default enrollment process is to match the percentage of non-choosers assigned to each plan with the percentage of all eligibles that voluntarily select the plan. In Florida and Massachusetts, enrollment in HMOs is voluntary; there, the percentage of non-choosers assigned to the HMO option is equal to the percentage that choose this option, and the distribution of these assignees to particular plans is likewise based on the distribution among voluntary enrollees. The first five of these methods are based on the assumption that those who fail to choose a plan are relatively healthy, and thus will not use high levels of services. Therefore, plans would be expected to compete, whether through price competition, enhanced levels of quality, or both, to receive preference in enrolling this population. The last approach described is based on the assumption that the auto-assigned population represents a random sample of Medicaid eligibles in the state; this population would not be expected to use services differently than a plan's voluntary enrollees. Continued research is needed to address which, if either, of these assumptions is correct. Health Systems Research, Inc. Chapter II Page 38 H. Quality Assurance Provisions Related to Capitation Rates Each state's contract with managed care plans outlines a wide range of quality assurance and improvement efforts, including the use of practice guidelines, member satisfaction surveys, HEDIS measures, and internal quality improvement systems. A number of these efforts relate directly to the development and payment of capitation rates, and are detailed below. Several states require plans to submit monthly or quarterly encounter data. These data, in addition to their use in calculating quality indicators, may be used to assess the adequacy of current rates and to calculate rates or rate guidelines for future years. Few states currently use encounter data for this purpose, however; Arizona is the only one that has established the capacity to collect consistent encounter data and to use these data as the primary data source for each bid year's rate guidelines, and Rhode Island uses encounter data as one of several sources of information on which to base rate guidelines. Other states, such as Delaware, Minnesota, Missouri, and Oregon have expressed a strong interest in using encounter data in this fashion but have been unsuccessful thus far in gathering consistent, high-quality data from managed care plans. A second approach to relating quality improvement efforts to capitation payments is to tie some portion of these payments to the achievement of specific objectives, such as immunization or EPSDT screening rates. Three states, Georgia, Tennessee, and Texas, identified specific efforts in this area, as described below. Georgia. In developing their state's Medicaid managed care initiative, Georgia officials were concerned that the program produce improvements in the state's traditionally low rate of EPSDT screening and immunization. Thus, to provide an incentive for plans to improve these rates (and to make the program more attractive to state legislators), the state included in its contracts the goal of screening 80 percent of children enrolled in EPSDT and immunizing 90 percent of children by age two. If plans did not screen or immunize at least 80 percent of children, they were to refund a portion of their capitation rates to the state Medicaid agency, according to the following schedule: I If screening rates are between 60 percent and 80 percent, the refund is to equal $10 per member under age 21 times the average lengths of enrollment in years of members under 21; Health Systems Research, Inc. Chapter II Page 39 - If screening rates are between 50 percent and 60 percent, the refund is to equal $20 per member under 21 per year of enrollment; - If screening rates are between 40 percent and 50 percent, the refund is to equal $30 per member under 21 per year of enrollment; - If screening rates are above the state average but below 40 percent, the refund is to equal $50 per member under 21 per year of enrollment; and - Regardless of the above standards, if screening rates are below the state average for children in fee-for-service or PCCM Medicaid arrangements, the refund is to equal $100 per member under 21 per year of enrollment. Thus, even if the state succeeds in improving its screening rates in its fee-for- service programs, the HMOs will continue to be required to equal or exceed these rates. Plans are exempt from these refund requirements for the first year of their contracts; thus, the results of this effort are not yet available. Texas. The Texas Medicaid agency has set standards for HMOs in a range of preventive health activities, including EPSDT screens, prenatal care, immunization, and adult preventive health screening. For each of 13 specific objectives, the state withholds $2 per member per month from the capitation payment for the appropriate eligibility group. These amounts are repaid to plans if they achieve each objective in each year. Examples of the specific objectives include: - Ninety percent of children under age one and ages one to two will receive at least one comprehensive EPSDT screen; - Seventy percent of children aged two through 20 years will receive at least one EPSDT screen; - Children under age one will receive an average of 3.8 EPSDT screens in their first year; - Ninety percent of children under age one and ages one to two will be fully immunized, as appropriate for their age; - Sixteen percent of adults age 21 and older will receive an annual health examination; - Sixty percent of pregnant women will receive a minimum of ten prenatal visits; and - Sixty percent of women who are pregnant at the time of enrollment in the plan will receive an initial prenatal exam within two weeks of initial enrollment. Again, since these contracts have only become effective this year, no results of this effort are yet available. Health Systems Research, Inc. Chapter II Page 40 Tennessee. Under the TennCare program, the state withholds 10 percent of each plan's monthly capitation rate, to be paid monthly based on the plan's ability to meet the requirements of the TennCare contract, including its quality assurance requirements. If the plan does not meet the contract's standards, the funds will be withheld each month until the deficiencies have been corrected. If a deficiency is not corrected within six months, the amount withheld is forfeited. Health Systems Research, Inc. Chapter II Page 41 CHAPTER III Conclusions and Future Directions In developing their strategies for establishing capitation rates, states must balance the competing interests of controlling Medicaid program costs (and, ideally, achieving savings) and attracting plans to the managed care initiative. While most states reported that their rate- development methodologies allowed them to meet both goals successfully, several have had difficulty striking an appropriate balance. Georgia, for example, has so far found only two plans that have agreed to accept the state's rates; Rhode Island was forced to raise its rate guidelines to meet what plans considered to be their minimum financial requirements; and Massachusetts officials report that they struggle each year to negotiate a rate with the plans in the state, as the state's providers require reimbursement rates from plans that are equal to those they receive for their commercially-insured patients. Another overriding issue of concern in many states is the need to identify appropriate sources of data on which to base capitation rates. As discussed throughout this report, most study states are currently using claims data from their fee-for-service Medicaid programs as the basis for their rate calculations. These databases may be drawn from previous years' claims or from claims in counties in which capitated managed care has not yet been introduced. State officials reported several disadvantages associated with the use of fee-for-service databases for this purpose: Using fee-for-service claims as the basis for capitation rates may preserve inequities and inefficiencies found in the fee-for-service system. For example, geographic areas and populations that are underserved in fee-for-service systems will be under-represented in the claims database; thus, the rates calculated will be based on the provision of low levels of service to these regions or groups. As described earlier, several states adjust their capitation rates in each service category to reflect the changes in utilization expected under managed care, and Minnesota requires that rates in rural counties be at least 85 percent of those in the Twin Cities metropolitan area. However, no efforts were found to Health Systems Research, Inc. Chapter III Page 42 systematically identify and adjust for traditionally low rates of utilization in specific populations. Several states have begun using primary care case management programs as an option to or in place of their traditional fee-for-service systems. In these cases, fee-for-service data will reflect the utilization patterns found in these PCCM systems, which may serve to control utilization somewhat through the use of the primary care provider as a case manager. Thus, if rates based on these programs' experience are further discounted, the final rates may be unrealistically low. Finally, as capitated managed care programs expand, new fee-for-service data will no longer be available. Thus, states will be required to use either outdated claims data or data from counties that are not served by the managed care program. The utilization patterns in these counties, however, may not be comparable to that in the counties in which capitated programs operate. This was reported to be a problem in Minnesota, where fee-for-service data will soon be available from only the most remote rural counties, and will thus be inapplicable to the health care environment in the rest of the state. To address these concerns, several states hope to use encounter data submitted by managed care plans as the basis for future rates. One state, Arizona, has established enough experience to routinely collect and base the rate calculation for each new bidding process on plans' encounter data. California is taking steps in this direction, using encounter data from one county with extensive and well-documented capitation experience as the basis for rate calculations for the 12 pilot counties in the Two-Plan Model. However, as the state must still comply with the UPL requirements of the state's 1915(b) waiver, the rates calculated must be adjusted to fall below the fee-for-service equivalent rates in the pilot counties. Finally, Rhode Island uses encounter data from plans as one of several sources of data for its actuarial database. Officials of several other states expressed interest in using encounter data for this purpose but have experienced difficulty collecting consistent, high-quality data from plans. Officials of several states indicated that they are planning or considering additional changes in their rate development methodologies. Several states that currently set capitation rates are considering instituting bidding and negotiation systems, for several reasons: In one state, an official felt that a bidding process would provide external validation of the rates developed by the state, thus assisting state actuaries in developing realistic rate guidelines in the future. Health Systems Research, Inc. Chapter III Page 43 In two states that currently contract with all plans that accept the state's rates, officials reported that competitive bidding processes may be used to reduce the total number of plans participating in the managed care program. In one predominantly rural state, this was described as a particularly important issue; a competitive process may be needed in rural areas that may not be able to support more than one managed care plan. In another state, officials have expressed interest in using competitive bidding to reduce the total number of contracts the Medicaid agency must manage. Two states expressed interest in competitive bidding processes because of their potential to reduce capitation rates. In addition to changes in their overall rate development methodology, several states are considering making changes to the rate cell structure within which current rates and rate guidelines are calculated. Two states that currently develop a single rate or rate guideline for each eligibility category are considering using age- and sex-specific rates for their greater accuracy and specificity, while two other states that use age and sex cells are considering consolidating these rates for the improvement in administrative simplicity. Overall, the survey demonstrated the wide variety of systems currently in use to establish capitation rates and negotiate contracts in Medicaid managed care systems. No single methodology emerged as clearly the most successful, and officials of the study states reported a variety of advantages and disadvantages of each method used. Moreover, states are continuing to modify their approaches to Medicaid managed care contracting, by modifying the methods used to develop rates for AFDC and related populations as well as by developing methods for setting specific rates for special populations. Health Systems Research, Inc. Chapter III Page 44 Attachment A: Interview Protocol Health Systems Research, Inc. Kaiser Capitation Survey State: Protocol Initial Contact: Phone Number: I. Introduction Hello. My name is . I'm with Health Systems Research, Inc., a health policy consulting firm in Washington, D.C. We're working with the Kaiser Family Foundation to conduct a study of how states contract with managed care plans for their Medicaid populations, with an emphasis on aspects of the bidding and contracting process that influence capitation rates. Because this is a pilot study, we are focusing our efforts on a select number of states and their managed care activities covering the AFDC and AFDC-related populations. We have selected 15 states for the study, including , based on their location, experience with Medicaid managed care, and penetration of managed care within the Medicaid and AFDC populations. The survey will take approximately 30 minutes; it will address the benefits included in the capitation rate for AFDC and related populations and the methodologies used to calculate these capitation rates. Are you the appropriate person to respond to this survey or is there someone more knowledgeable about these issues with whom I should speak? Contact: Title: Phone number: [if this is the right person]: Is this a good time or may I schedule another time? Time: II. Background [before interview, fill in this section based on HSR's information, and confirm] 1. First, I'd like to confirm some basic information about Medicaid managed care in [state]. As I understand it, your managed care initiatives include: 1 2. The eligibility categories included in capitated managed care include: 3. Which eligibility categories are specifically excluded from capitated systems? 4. We are focusing on AFDC and related eligibility categories, including those who receive cash payments, those who are in AFDC non-cash categories (such as transitional Medicaid), and the expanded eligibility categories for pregnant women and children. The state has been enrolling AFDC and related eligibles in managed care since . Enrollment for these populations is (voluntary/mandatory). 5. Within the AFDC and related population, are there any subpopulations for whom separate capitation rates are developed, such as pregnant women with HIV? 6. The state is currently providing managed care services for the AFDC and related populations under waivers. 7. Are any of these programs currently operating statewide? If not, which geographic areas are currently included? Is the population being phased in? 8. How are geographic areas defined for the purposes of bidding and contracting with managed care plans? [if there is more than one capitated program for AFDC]: I'd like to ask you to choose the most important or largest of the capitated programs (with respect to the AFDC population) to focus on (one that provides at least basic primary and acute care services for AFDC-eligible and related categories of children and families). We'll come back to the others after we've explored that one in detail. Program being discussed: 2 III. Benefits Now I'd like to turn to the benefits included in the capitation rate for AFDC and related eligibles. 1. Which required Medicaid benefits are included in the base capitation rate? 2. Which required benefits are specifically excluded from the capitation rate? (Use the list below to record answers; if included, X if excluded. If they don't mention them, confirm that services below the horizontal line are included.) Required Benefits Primary and preventive care Outpatient substance abuse treatment Specialty care Inpatient substance abuse treatment Inpatient hospital care Other: Outpatient hospital services Family planning Immunizations Laboratory services EPSDT services EPSDT outreach and informing functions EPSDT treatment services (outside the State Plan) Home health services Outpatient mental health services Inpatient mental health services 3 3. What optional benefits are included in the capitation rate? Optional Benefits Prescription drugs Adult Dental Durable medical equipment Physical/occupational/speech therapy Enhanced/support services for pregnant women (not covered ) Enhanced/support services for high-risk infants (not covered ) Enhanced/support services for children at risk (not covered ) Transportation Translation Part H Early Intervention Services Other: 4. Are any subcategories of service carved out of the base capitation rate for specific subpopulations (e.g., specialty services for children enrolled in the state's children with special health care needs program)? 5. Are plans paid a supplemental premium or fee-for-service for any specific services (such as transportation services)? 4 6. Does the state impose limits on any services (e.g., number of outpatient mental health visits or physical therapy sessions per year?) 7. Are plans permitted to set their own limits on the use of particular services? 8. Are specific definitions of any services provided to plans, such as for case management? 9. Are plans required to provide any additional, specific health promotion and disease prevention services, such as smoking cessation or nutrition classes? Encouraged? 10. Are plans permitted to charge copayments for any services? Do any do so? For what services? IV. Development of Capitation Rates Now I'd like to turn to how you arrived at the capitation rates you pay to plans for these services. First, can you briefly define the type of process you use for developing capitation rates? When was the last time you went through the contracting process with managed care plans? For the remainder of this discussion, I'd like to focus on that contracting process, beginning with the information contained in your RFP. 5 The RFP 1. In your RFP, how much flexibility did you give plans in developing proposed capitation rates? Did you specify an upper limit? A lower limit? A target rate? 2. For the AFDC and related populations, how many different rate cells were plans asked to present? How did you define the populations to be covered? What factors are used to define the populations for which individual rates are developed? Age? (groups: ) Sex Eligibility category (Categories: ) Geography (areas: ) Other: 3. In addition to these factors, is the base capitation rate risk-adjusted based on the health status or utilization experience of the enrolled population? For example, are plans paid an enhanced rate for the care of high-risk pregnant women or children with high-cost conditions? (Again, we're referring to the AFDC population only, not SSI eligibles.) If yes, how is this implemented? Are plans paid an adjusted rate up front based on a risk screen, or are plans paid retrospectively, based on utilization of services? Or do you use both mechanisms? 6 [if adjustment is based on a risk screen]: What risk factors are included in this screen? Could you send me a copy of the screening form? [if adjustment is based on utilization]: What level of utilization triggers an increase in the rate? How is the risk-adjusted rate calculated? 4. What information did you provide to plans to help them develop proposed rates? How much detail was provided? A target rate for each cell, such as 95% of AAPCC(Average Annual Per Capita Cost or Historical Cost) Information about historical utilization and cost of each service, by eligibility category Was this information provided on paper or on a data tape that the plans can manipulate? 5. [if estimates of 95% AAPCC or other target are given to bidders]: What was the base year used to calculate this rate? Did you consult with an actuary in the development of the target rate? Did you use the actuarial estimates? What factor was used to trend the historical rate forward? Were all services included in the historical claims used to calculate the rate? Were any adjustments made to the original calculation (for varying costs in different regions of the state, for example)? 7 6. What information were plans required to present to support their proposed rates? Evaluation of Bids 1. How are plans' bids judged? Do you set out criteria on which proposals are scored, with cost as one aspect of the aggregate score? or do you score bids on all non-cost (quality) items, then evaluate the cost aspect of those that meet a certain standard? [If cost is one aspect of the proposal's score] How are the various factors weighted? How are these factors quantified? [If proposals are scored first on all non-cost items] What are the major evaluation criteria? Briefly, how are these evaluated and quantified? 2. How are the cost proposals evaluated? Are proposals eliminated if their proposed rates are too far above or below a pre-determined standard? What are these limits and how are they set? Are these comparisons made within each individual rate cell, or for the overall average rate only? 8 Negotiation and Contracting 1. Do you then negotiate the rate with the plans? What information is given to the plans to prepare for this negotiation? Do you give them a target, or just tell them their rate is too high or too low? If you give them information, how much detail do you provide? How much flexibility is there in the negotiation process? Do you have predetermined upper and lower bounds for the final capitation rate? What are these limits and how are they determined? 2. At the conclusion of the negotiation process, what was the range of rates paid to plans? Are the final rates public information? That is, can plans find out what rates other plans in the same service area receive? 3. Do you limit the number of plans that you contract with in each county or region? What is that limit and how is it set? 4. How is each plan's enrollment capacity determined? Is there a lock-in period for enrollees? 9 5. Do you have any provision for assigning enrollees based on the range of rates? 9. What is the time period of your contracts with managed care plans? How are rates adjusted from year to year? 10. When the contract comes up for re-bid, will new rates be determined, or will existing contractors be permitted to continue to use their existing rates, with adjustments for inflation? Sharing of Risk 1. Do you set risk corridors, in which the state shares in savings or shares risk within certain bounds? If so, what are those bounds and how are they set (e.g., as a percentage of the total capitation rate)? 2. Did the state develop stop-loss mechanisms or reinsurance requirements? Is this reinsurance provided by the state, or are plans required to obtain reinsurance on the private market? Is the cost of reinsurance included in the capitation rate? At what level of expenditures are these systems required to "kick in"? 3. Is any additional compensation offered to plans for reaching service targets (such as 10 EPSDT screening goals)? [if more than one capitated system for AFDC]: Now let's turn to the next statewide capitated program for AFDC and related populations. What is this program called? How does it differ from the one we just discussed? [return to page 3] V. Lessons Learned 1. The process you just described is the one you used the last time you put contracts with managed care plans out for bid. Has this process changed from the one you used for earlier bids? If so, can you describe briefly how it has changed? Why did you implement a new process? What problems did you encounter with the original system? Did the new process address these problems successfully? 2. What would you like to change about your current process for determining capitation rates? 11 What would you need to do to make these changes? 3. What do you see as the major advantages of the process you currently use? VI. Quality Monitoring Systems Next I'd like to ask a few questions about the systems you have in place to monitor the use of Medicaid funds paid through these capitation rates and the quality of care provided by plans. 1. What types of information do you collect from plans to monitor the quality of care? Do you use HEDIS? other performance measures? What other information is collected (encounter data? cost data?) What supporting documentation do you require plans to submit? 2. Do you use this information to assess the adequacy of the rates paid to plans? 3. Do you plan to use this information to affect the calculation of rates for future years? 12 4. Who is responsible for monitoring the solvency of plans that contract with Medicaid? What information does the Medicaid agency provide to assist in this assessment? Insurance commissioner? VII. Wrap-up Thank you very much for your time. As we proceed with our surveys and analyses, we may want to call you back to clarify some of the information you've given me. Is that ok? Also, would you be able to send me copies of supporting documents, including: The RFP you sent out, with any supporting or background information Your general contract with managed care plans Your waiver application 13 The Henry J. Kaiser Family Foundation Washington Office: 2400 Sand Hill Road 1450 G Street, N.W., Suite 250 Menlo Park, CA 94025 Washington, DC 20005 650-854-9400 Facsimile 650-854-4800 202-347-5270 Facsimile 202-347-5274 Request for Publications: 800-656-4533 http://www.kff.org