oversight

Medicaid: States' Use of Managed Care

Published by the Government Accountability Office on 2012-08-17.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

United States Government Accountability Office
Washington, DC 20548



           August 17, 2012

           The Honorable John D. Rockefeller, IV
           Chairman
           Subcommittee on Health Care
           Committee on Finance
           United States Senate

           The Honorable Henry A. Waxman
           Ranking Member
           Committee on Energy and Commerce
           House of Representatives

           Subject: Medicaid: States’ Use of Managed Care

           The Medicaid program, a joint federal-state program that finances health insurance
           coverage for certain categories of low-income individuals, is an important source of
           health care coverage for about 67 million beneficiaries. As Medicaid enrollment and
           spending have increased significantly over the past decade, so too has states’ use
           of managed care to provide services to Medicaid beneficiaries, and nearly all states
           enroll some Medicaid beneficiaries in a form of managed care. Within some general
           requirements set out by the Centers for Medicare & Medicaid Services (CMS), the
           federal agency responsible for overseeing the Medicaid program, states have broad
           flexibility to implement Medicaid managed care programs. As a result, states vary
           widely in terms of the scope of services they provide and the populations they enroll
           in managed care. For example, while states commonly contract with managed care
           organizations (MCO) to provide the full range of covered Medicaid services to
           certain enrollees, 1 they also frequently rely on other arrangements, such as limited
           benefit plans, 2 which provide a limited set of services, such as dental care or
           behavioral health services, or primary care case management (PCCM) programs, in
           which enrollees are assigned a primary care provider (PCP) who is responsible for
           providing primary care services and for coordinating other needed health care




           1
            States pay MCOs a set, or capitated, per member per month fee to provide enrollees access to
           contracted services and coordination of care.
           2
            Some states enroll Medicaid beneficiaries into limited benefit plans, which generally are paid on a
           prepaid basis for providing a limited set of covered services, such as dental care, behavioral health
           care, and transportation, to beneficiaries.



                                                                      GAO-12-872R Medicaid Managed Care
services. 3 States also vary in their use of managed care for other reasons, such as
differences in the availability of certain providers or the concentration of program
beneficiaries that live in urban or rural areas.

The Patient Protection and Affordable Care Act (PPACA) of 2010 requires that all
states expand eligibility for Medicaid to nonelderly individuals whose income does
not exceed 133 percent of the federal poverty level (FPL); 4 this expansion is
estimated to result in the enrollment of an additional 7 million individuals in 2014. 5 As
initially set forth in PPACA, states that did not fully implement this Medicaid
expansion faced the potential loss of all federal Medicaid matching funds, including
for the population already covered under existing program rules. However, the U.S.
Supreme Court has ruled that states that choose not to expand Medicaid eligibility to
these newly eligible individuals will forgo only the federal matching funds associated
with such expanded coverage. 6 States that choose to provide Medicaid services to
newly eligible individuals may do so through managed care arrangements. 7

Because of your interest in the potential increase in Medicaid managed care
enrollment and related implications, you asked us to describe states’ use of
Medicaid managed care, including the type of managed care arrangements they
have in place, and their enrollment of populations with complex health care needs. 8
Understanding how states use Medicaid managed care—and related similarities and
differences among them—may be informative as states consider expanding their
use of managed care to new geographic areas or new populations, such as disabled
beneficiaries who traditionally have more complex health care needs. This report
examines variation in states’ use of Medicaid managed care, and identifies groups of

3
 GAO has historically described PCCM programs as a predominantly fee-for-service arrangement
because most services provided by participating PCPs are reimbursed on a fee-for-service basis.
Under a PCCM system, states pay participating PCPs a monthly, per person case management fee
for coordinating enrollee health care services, and separately reimburse them on a fee-for-service
basis for specific health care services they provide. For purposes of this report, however, we include
PCCM programs in the broader discussion of managed care arrangements, which is consistent with
CMS’s current practice. In addition, CMS officials noted that the agency is thinking more broadly
about how PCCM authority can be used in the future in a non-managed care delivery system.
4
 Patient Protection and Affordable Care Act of 2010 (PPACA), Pub. L. No. 111-148, 124 Stat. 119, as
amended by the Health Care and Education Reconciliation Act of 2010 (HCERA), Pub. L. No. 111-
152, 124 Stat. 1029. For purposes of this report, references to PPACA include the amendments
made by HCERA.
5
 Pub. L. No. 111-148, § 2001, 124 Stat. 271. The 7 million estimate includes new enrollment in both
Medicaid and the Children’s Health Insurance Program. See Congressional Budget Office, Estimates
for the Insurance Coverage Provisions of the Affordable Care Act Updated for the Recent Supreme
Court Decision (Washington, D.C.: July 2012).
6
 See National Federation of Independent Business, et al., vs. Sebelius, Sec. of Health and Human
Services, et al., No. 11-393 (U.S. June 28, 2012).
7
 States will receive an increased federal match for newly eligible individuals at 100 percent for 2014
through 2016, 95 percent in 2017, 94 percent in 2018, 93 percent in 2019, and 90 percent in 2020
and beyond. Pub. L. No. 111-148, §§ 2001(a)(3), 10201(c), 124 Stat. 272, 918, as amended by
Pub. L. No. 111-152, § 1201, 124 Stat.1051.
8
 Our definition of “Medicaid beneficiaries with complex health care needs” includes beneficiaries who
were aged, blind/disabled, medically needy, or dually eligible for Medicare and Medicaid. We
excluded from our definition dual eligibles who only received Medicare cost-sharing assistance
through Medicaid.



2                                                           GAO-12-872R Medicaid Managed Care
states that share similarities, such as program enrollment composition and general
market characteristics.

To examine variation in states’ use of Medicaid managed care, we reviewed multiple
data sources, such as CMS’s Medicaid Statistical Information System (MSIS) and
the Census Bureau’s American Community Survey (ACS), and ultimately identified
12 indicators that were informative in understanding the context in which states use
Medicaid managed care. 9 The indicators are grouped into two broad categories:
(1) population-based characteristics, such as state-reported enrollment in MCOs and
PCCM programs 10 and the degree of potential Medicaid expansion that could occur
in 2014; 11 and (2) state market and other characteristics, such as the health
maintenance organization (HMO) penetration rate, and the concentration of low-
income individuals who lived in urban areas. 12 We excluded other indicators, such as
states’ regulatory environment and use of limited benefit managed care plans due to
the lack of available or reliable data. Specifically, we excluded data on oversight
activities because they were not available in a format that was suitable for our
analysis, and enrollment in limited benefit plans because of inconsistencies in state-
reported data.

We then conducted a cluster analysis, a statistical method that assessed these
indicators simultaneously in an effort to cluster states into groups, which were as
similar as possible on the indicators within groups and as different as possible
among the groups. Cluster analysis is a technique that allows us to focus on broad,
shared patterns among states and can yield insights that are difficult to discern just
by looking at simple comparisons of data across states. States that are similar with
respect to multiple indicators may be able to gain insights from each other in terms
of administering or expanding their Medicaid managed care programs. We also
interviewed officials from CMS and other national policy experts, including officials

9
 The 12 indicators were: (1) the percentage of Medicaid beneficiaries enrolled in MCOs; (2) the
percentage of Medicaid beneficiaries with complex health care needs enrolled in MCOs; (3) the
percentage of Medicaid beneficiaries enrolled in PCCM programs; (4) the percentage of Medicaid
beneficiaries with complex health care needs enrolled in PCCM programs; (5) the Medicaid
Expansion Index, which is the degree of potential Medicaid expansion; (6) the concentration of low-
income individuals that lived in urban areas; (7) the HMO penetration rate; (8) the commercial HMO
Market Competition Index; (9) the number of MCOs per 100,000 Medicaid beneficiaries; (10) the
Primary Care Capacity Index; (11) the allowable PCPs in MCOs; and (12) the allowable PCPs in
PCCM programs.
10
 Enrollment data are derived from state-reported data to CMS’s MSIS and provide detailed
enrollment for the various managed care arrangements states have in place.
11
  The Medicaid Expansion Index is derived from projections of the number of individuals considered
potentially eligible for Medicaid in 2014 as a result of PPACA’s expansion of eligibility in relation to the
number of low-income individuals in the state. For more details, see L. Ku, K. Jones, P. Shin, B.
Bruen, and K. Hayes, “The State’s Next Challenge—Securing Primary Care for Expanded Medicaid
Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI:
10.1056/NEJMp1011623. This measure was developed prior to the Supreme Court decision and
assumes expanded participation by all states; however, the extent to which states will implement
PPACA’s eligibility expansion is uncertain at this time. Its purpose as a measure is not to assume
states’ actions with regard to expanding Medicaid, but to provide a relative indicator of the extent of
potentially eligible individuals within a state.
12
  Historically, there have been differences in state implementation of Medicaid managed care in
urban and rural areas; states have been more likely to contract with MCOs to provide care in urban
areas.



3                                                             GAO-12-872R Medicaid Managed Care
from the Medicaid and CHIP Payment and Access Commission and the National
Association of Medicaid Directors, and reviewed published reports and surveys
related to states’ use of Medicaid managed care. 13 (See enc. I for more information
on our scope and methodology, and enc. III for a detailed description of the
indicators we examined.)

To determine the reliability of data sources we identified, we reviewed related
documentation and conducted electronic testing for missing data, outliers, and
apparent errors, and determined that the data were sufficiently reliable for our
purposes. We conducted this performance audit from March 2011 through
August 2012 in accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain sufficient,
appropriate evidence to provide a reasonable basis for our findings and conclusions
based on our audit objectives. We believe that the evidence obtained provides a
reasonable basis for our findings and conclusions based on our audit objectives. We
completed our field work prior to the June 28, 2012 decision by the Supreme Court
on certain aspects of PPACA, including the Medicaid expansion provision.

In summary, we identified four groups of states that differed in their use of Medicaid
managed care on the basis of the 12 indicators we included in our analysis. A
handful of these indicators—namely Medicaid enrollment in MCOs and PCCM
programs, HMO penetration rates, and the concentration of low-income individuals
that lived in urban areas—had significant influence on how states grouped. In
contrast, within the four groups, considerable variation existed among the other
indicators we examined, such as states’ primary care capacity and commercial HMO
market index. For labeling purposes, we typically describe the four groups on the
basis of states’ enrollment of Medicaid beneficiaries in MCOs and PCCM
programs—generally the predominant similarity among the states within each group
(see fig. 1.):

•     Group 1 states were PCCM predominant, enrolling a high percentage of
      beneficiaries in PCCM programs, but typically not in MCOs;

•     Group 2 states typically enrolled beneficiaries in both MCOs and PCCM
      programs;

•     Group 3 states were MCO predominant, enrolling a high percentage of
      beneficiaries in MCOs, but typically not in PCCM programs; and

•     Group 4 states were considered “other” states in that although their enrollment of
      beneficiaries was similar to Group 3, they were outliers on other indicators, which
      differentiated them from states in the other groups we identified. 14
Enclosure II provides additional information on these groups of states, and
enclosure III provides state-specific data related to each of the indicators.
13
    We use “states” to refer to the 50 states and the District of Columbia for the purposes of this report.
14
  In addition, the similarity identified among states in Group 4 was weaker than the similarities
identified among states in Groups 1, 2, and 3.



4                                                              GAO-12-872R Medicaid Managed Care
Figure 1: Summary of Selected Indicators by State Groups




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file, U.S. Census Bureau 2005-2009 American Community Survey, and Kaiser Family Foundation data.
Some states in each group were exceptions to the general group descriptions presented in this figure. We use “states” to refer
to the 50 states and the District of Columbia.


The cluster analysis results provide perspective on how states have implemented
Medicaid managed care and highlight strong similarities shared among states within
each of the groups, particularly with regard to MCO and PCCM enrollment. States
within each of these groups could look to one another as a resource as they
consider expanding their Medicaid managed care programs.

The results also provide specific information about challenges states may face in
expanding their use of Medicaid managed care. For example, each of the groups
emerging from our analysis included states that may face greater than average
Medicaid program expansions in 2014 if they fully implement PPACA’s eligibility
expansion, and it is likely that many of these states will look to managed care to
provide services to their newly eligible population. Specifically, 10 of the 12 states
with the greatest potential Medicaid expansion are in Groups 1 and 2—states with


5                                                                        GAO-12-872R Medicaid Managed Care
high enrollment in PCCM programs only or using a mix of MCOs and PCCM
programs. However, 8 of these 10 states have a below average primary care
capacity, and 9 of the 10 states had a comparatively small concentration of low-
income individuals that lived in urban areas or a low HMO penetration rate, which
may affect the states’ capacity to expand their managed care programs to serve
additional beneficiaries. For example, these states may face challenges attracting
MCOs that may have concerns about the availability of adequate provider networks
or the sufficiency of enrollment, and thus may not have a strong business case for
entering the state’s Medicaid managed care market. Similarly, states with low
primary care capacity may not have enough providers to expand their PCCM
programs further, or MCOs that the state contracts with may have challenges
building an adequate network of PCPs. Therefore, in determining whether to
implement or expand the use of Medicaid managed care and related challenges,
these states will need to consider these indicators, as well as other contextual
factors that may affect their capacity to do so, and may look to similarly situated
states for guidance.

Despite the robustness of our analysis, it provides an incomplete picture because
data on additional indicators that affect states’ implementation of Medicaid managed
care were not available or were unreliable. For example, data on states’ Medicaid
program oversight capacity and activities could provide a more complete picture of
states’ Medicaid managed care programs and related challenges, and could provide
insight on resources and expertise they may need to expand their managed care
programs. Similarly, reliable enrollment data for limited benefit plans would provide a
more comprehensive picture of states’ use of Medicaid managed care. A cluster
analysis that includes these data would offer even more robust groupings of states,
which could be more useful for states that are considering Medicaid managed care
expansions. Ensuring the availability of more complete and reliable data and
conducting research on additional indicators will be important to developing a more
comprehensive picture of how states use Medicaid managed care.

Agency Comments

We provided a draft of this report to the Department of Health and Human Services
(HHS) for comment. HHS responded that it did not have any comments on the draft
report.
                                        –––––
As we agreed with your offices, unless you publicly announce the contents of this
report earlier, we plan no further distribution of it until 30 days from its date. We are
sending copies of this report to the Secretary of Health and Human Services and
other interested parties. In addition, the report is available at no charge on the GAO
website at http://www.gao.gov.




6                                                   GAO-12-872R Medicaid Managed Care
If you or your staff have any questions about this report, please contact me at
(202) 512-7114 or yocomc@gao.gov. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page of this
report. GAO staff who made major contributions to this report are listed in
enclosure IV.



Carolyn L. Yocom
Director, Health Care
Enclosures – 4




7                                               GAO-12-872R Medicaid Managed Care
Enclosure I
                                    Scope and Methodology

To examine how states used managed care to provide services to their Medicaid
beneficiaries, we conducted a cluster analysis, a statistical method that allowed us to
identify groups of states that were similar across multiple characteristics
simultaneously. As a first step, we identified the following data sources to collect
specific information related to states’ use of managed care and other related
characteristics:
•   The Centers for Medicare & Medicaid Services’ (CMS) Medicaid Statistical
    Information System (MSIS) Annual Person Summary File (APS): The MSIS APS
    file contains individual-level demographic, enrollment, and service utilization data
    summarized for each beneficiary at an annual level from quarterly MSIS files
    submitted by states. 1

•   CMS’s Medicaid Managed Care Data Collection System (MMCDCS): The
    MMCDCS includes state-reported information on states’ Medicaid managed care
    programs, including enrollment by type of managed care plan; program
    characteristics, such as the types of providers states permit to act as a primary
    care provider (PCP) in their Medicaid managed care programs; and state
    program management activities. 2

•   Census Bureau’s American Community Survey (ACS): The ACS provides data
    annually on population demographics, income, health insurance, education,
    employment, and other characteristics. We used information from the ACS to
    determine the concentration of low-income individuals that lived in urban areas in
    each state. 3



1
 At the time of our analysis, 2008 was the most recent year for which MSIS APS data were available;
however, 2009 MSIS APS data are now available. Given the extensive time and effort required to
identify and correct inconsistencies in the 2008 MSIS APS enrollment data, we opted to rely on the
2008 data rather than taking additional time to undertake similar data cleaning efforts with the 2009
data. State Medicaid agencies provide CMS with quarterly electronic files through MSIS that contain
data on: (1) persons covered by Medicaid, known as “eligible files;” and (2) adjudicated claims, known
as “paid claims files,” for medical services reimbursed by the Medicaid program. Each state’s eligible
file contains one record for each person covered by Medicaid for at least 1 day during the reporting
quarter; eligible records consist of demographic, eligibility, and monthly enrollment data. Paid claims
files contain information on medical service-related claims and capitation payments, but only include
expenditures that can be linked to a specific enrollee. The APS, however, summarizes the
demographic, eligibility, enrollment, utilization, and expenditure data for each person for an entire
fiscal year.
2
 At the time of our analysis, 2009 MMCDCS data were the most recent available. CMS indicated that
overall MMCDCS enrollment data are more reliable than MSIS enrollment data; however, MMCDCS
data do not provide enrollment on the basis of Medicaid eligibility category, which was integral to our
analysis.
3
 Low-income individuals are defined as those individuals with incomes below 125 percent of the
federal poverty level (FPL). Urban areas comprise areas that consist of a central place(s), have a
minimum population density of 1,000 people per square mile, and have an overall minimum
population of 50,000 people. It also includes adjacent areas that have lower population density but
are linked to the more densely settled area and have a population of at least 2,500 people, but fewer
than 50,000.



8                                                            GAO-12-872R Medicaid Managed Care
Enclosure I

•   Kaiser Family Foundation data sources on states: The foundation’s website
    included information compiled by another organization on health maintenance
    organization (HMO) penetration rates, and the results of the foundation’s 50-state
    survey on states’ use and future planned uses of Medicaid managed care. 4

•   American Medical Association (AMA) data: A recent AMA report on competition
    in the health insurance industry included measures of state-level competition
    among commercial HMO plans based on data from another organization.

•   Leighton Ku and colleagues’ data published in the New England Journal of
    Medicine: A 2010 journal article on states’ capacity to meet expanded demand
    for health care services if states fully implement the Medicaid eligibility
    expansions under PPACA included measures of state-level primary care capacity
    and potential increases in Medicaid enrollment. 5

From these sources, we identified 12 indicators in 2 general categories of states’ use
of Medicaid managed care: (1) population-based characteristics, such as enrollment
in managed care organizations (MCO) and primary care case management (PCCM)
programs; and (2) state market characteristics. (Table 1 and enc. III provide a
detailed description of the indicators we examined.)




4
 The Kaiser Family Foundation recently conducted a survey of state Medicaid managed care
programs to collect information on states’ use of managed care and their future plans to use or
expand the use of managed care. For more details, see K. Gifford, V. K. Smith, D. Snipes, and
J. Paradise, A Profile of Medicaid Managed Care Programs in 2010: Findings from a 50-State Survey
(Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on Medicaid and the Uninsured,
September 2011).
5
 The Primary Care Capacity Index is based on measures of states’ primary care workforce in relation
to the state’s population. For purposes of this index, the primary care workers include internists,
family or general practitioners, pediatricians, obstetricians/gynecologists, nurse practitioners, and
physician assistants. It also accounts for the number of patients seen at federally qualified health
centers. The Medicaid Expansion Index is derived from projections of the number of individuals
considered potentially eligible for Medicaid in 2014 as a result of the Patient Protection and Affordable
Care Act’s (PPACA) expansion of eligibility in relation to the number of low-income individuals in the
state. For more details, see L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next
Challenge—Securing Primary Care for Expanded Medicaid Populations,” New England Journal of
Medicine, vol. 364, no. 6 (2011), DOI: 10.1056/NEJMp1011623. This measure was developed prior to
the U.S. Supreme Court ruling that states may choose not to expand coverage under PPACA and
forgo only the federal matching funds associated with such expanded coverage, and assumes
expanded participation by all states. See National Federation of Independent Business, et al., vs.
Sebelius, Sec. of Health and Human Services, et al., No. 11-393 (U.S. June 28, 2012). However, the
extent to which states will fully implement PPACA’s eligibility expansion is uncertain at this time.



9                                                           GAO-12-872R Medicaid Managed Care
Enclosure I

Table 1: Description of Indicators Included In the Cluster Analysis

Indicator name         Description of indicator                Year of data       Source
Population-based Characteristics
Percent of Medicaid    Percentage of Medicaid beneficiaries Fiscal year 2008      Medicaid Statistical
beneficiaries          that were enrolled in an MCO and had                       Information System (MSIS)
enrolled in managed    a capitated payment made on their                          Annual Person Summary
                                          a
care organizations     behalf to an MCO.                                          (APS) file
(MCO)
Percent of Medicaid    Percentage of Medicaid beneficiaries    Fiscal year 2008   MSIS APS file
beneficiaries with     with complex health care needs that
complex health care    were enrolled in an MCO and had a
needs enrolled in      capitated payment made on their
                                         a,b
MCOs                   behalf to an MCO.
Percent of Medicaid    Percentage of Medicaid beneficiaries Fiscal year 2008      MSIS APS file
beneficiaries          that were enrolled in a PCCM program
enrolled in primary    and had a per member per month
care case              case management fee made on their
                                                  c
management             behalf to a PCCM provider.
(PCCM) programs
Percent of Medicaid    Percentage of Medicaid beneficiaries Fiscal year 2008      MSIS APS file
beneficiaries with     with complex health care needs who
complex health care    were enrolled in a PCCM program and
needs enrolled in      had a per member per month case
PCCM programs          management fee made on their behalf
                                           b,c
                       to a PCCM provider.
                                                                                                    e
Medicaid Expansion Index based on the number of                Calendar years     Ku et al., 2010
                                                                         d
Index              uninsured, nonelderly adults with           2007-2009
                   incomes below 138 percent of the
                   federal poverty level (FPL) who are
                   either currently eligible for Medicaid,
                   but not insured, or who could become
                   eligible for Medicaid if states fully
                   implement PPACA’s Medicaid
                                                        d
                   expansion requirements in 2014.
State Market and Other Characteristics
Concentration of       Percentage of the population earning    Calendar years     American Community
low-income             less than 125 percent of the FPL that   2005-2009          Survey
individuals in urban   lived in urban areas.
areas
Health maintenance Percentage of the total population          July 2010          Kaiser Family Foundation
organization (HMO) enrolled in an HMO plan, such as a
penetration rate   commercial HMO plan, or a Medicaid
                   or Medicare managed care plan.
Commercial HMO         A Herfindahl-Hirschman Index (HHI) of January 2009         American Medical
market competition     the competitiveness of the statewide                       Association (AMA)
index                  commercial HMO market. The index is
                       the sum of the squared market share
                                                    f
                       of all HMO plans in a state.
Number of MCOs         Number of MCOs that each state          October 2010       Kaiser Family Foundation
per 100,000            contracts with per 100,000 Medicaid
Medicaid               beneficiaries.
beneficiaries
                                                                                                    e
Primary Care           Index based on the number of primary Calendar years        Ku et al., 2010
                                                                      g
Capacity Index         care providers (PCP) and the number 2008-2009
                       of unduplicated patients seen at
                       federally qualified health centers in the
                              g
                       state.




10                                                               GAO-12-872R Medicaid Managed Care
Enclosure I

    Indicator name       Description of indicator                      Year of data           Source
    Allowable PCPs:      Average number of different types of          June 2009              Medicaid Managed Care
    MCO                  providers that a state allows MCOs to                                Data Collection System
                                                                                                         h
                         consider as PCPs.                                                    (MMCDCS)
                                                                                                          h
    Allowable PCPs:      Average number of different types of          June 2009              MMCDCS
    PCCM                 providers that a state allows to
                         participate as PCPs in its PCCM
                         program.
Source: GAO.
a
 In general, we considered individuals as enrolled in an MCO if they were reported as being enrolled in an MCO and had a
capitated payment made to an MCO on their behalf. For three states—Alabama, Idaho, and Utah—we assumed there was no
MCO enrollment in 2008 to address data reporting issues. Specifically, Alabama did not enroll Medicaid beneficiaries into
MCOs, but reported dually eligible enrollees for whom Medicaid pays for Medicare cost sharing as an MCO enrollee. Idaho and
Utah erroneously reported enrollment in MCOs; neither state enrolled Medicaid beneficiaries into MCOs in 2008.
b
 Our definition of “Medicaid beneficiaries with complex health care needs” includes beneficiaries who were aged,
blind/disabled, medically needy, or dually eligible for Medicare and Medicaid. We excluded dual eligibles who only received
Medicare cost-sharing assistance through Medicaid from our definition.
c
 In general, we considered individuals as enrolled in a PCCM program if they were reported as being enrolled in a PCCM
program and had a monthly case management fee paid on their behalf. Because of the way Colorado, Delaware, Maine,
Massachusetts, New York, South Dakota, and Utah reported PCCM case management payments, we relied only on the PCCM
enrollment data to determine the number of enrollees. To determine PCCM enrollment in Oklahoma, we relied on the limited
benefit plan enrollment data because Oklahoma typically reports those enrolled in its comprehensive PCCM program as
enrolled in a limited benefit plan instead.
d
 The Medicaid Expansion Index is based on two measures: (1) adults aged 19 to 64 who could become eligible, or who are
already eligible for Medicaid but not enrolled; and (2) an Urban Institute estimate of individuals who may become newly eligible
for Medicaid and may enroll. The number of individuals who would potentially become eligible for Medicaid under PPACA’s
eligibility expansion is standardized by the number of the individuals with incomes below 200 percent of FPL. This index was
calculated using data from calendar years 2007 to 2009 and assumes that all states will expand Medicaid eligibility.
e
L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing Primary Care for Expanded
Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI: 10.1056/NEJMp1011623.
f
The market share is calculated on the basis of the number of insured individuals enrolled in a single, commercial HMO plan
divided by the sum of all commercial HMO enrollment in a state, which is then multiplied by 100.
g
 The number of “primary care providers” included the number of internists, family/general practitioners, pediatricians,
obstetricians/gynecologists, 50 percent of the number of nurse practitioners and physician assistants, and the unduplicated
number of patients seen at federally qualified health centers (FQHC). To create this index, the number of primary care
providers was standardized by the state population, and the number of patients served by FQHCs was standardized by
number of people with incomes below 200 percent of the FPL. This index was based on the following data sources: estimates
of nonfederal physicians from the AMA; nurse practitioners in 2009 based on the Pearson Report; projected number of
physician assistants in December 2008 from the American Academy of Physician Assistants; number of patients served in
FQHCs in 2009 from the Health Resources and Services Administration, Bureau of Primary Health Care’s Uniform Data
System.
h
 Because of reporting issues in the 2009 MMCDCS for Wisconsin and Vermont, for our cluster analysis, we substituted data on
the number of allowable primary care types in the state’s MCO program(s) for Minnesota and the number of allowable types in
Vermont’s PCCM program with data collected by the Kaiser Family Foundation in a 50-state survey on Medicaid managed
care. See K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care Programs in 2010: Findings
from a 50-State Survey (Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on Medicaid and the Uninsured,
September 2011).


Limitations of Our Analysis

Our analysis captures state information with respect to the selected indicators
described above; however, several other factors that reflect or affect states’ use of
Medicaid managed care were not included in our analysis. For example, states vary
in their use of limited benefit plans, which provide a limited set of Medicaid benefits,
such as behavioral health and dental services, and are a key component of the
overall Medicaid managed care delivery system in some states. 6 Although the MSIS
6
 Previous research has shown that Medicaid beneficiaries sometimes experience challenges in
accessing behavioral health and dental care. For some populations, such as individuals with complex
health care needs, these access challenges can be especially problematic.



11                                                                        GAO-12-872R Medicaid Managed Care
Enclosure I

data included information on states’ enrollment in limited benefit plans, we identified
cases where state-reported data were incomplete or inconsistent, and thus
determined that the data were not reliable for our purposes. 7 Another factor we
considered, but ultimately excluded from our analysis, was the extent of states’
Medicaid managed care program management and oversight activities. According to
CMS officials, the agency’s most readily available data source on state Medicaid
managed care programs—the MMCDCS—contains descriptive and qualitative data
on state program management and oversight activities, which were not suitable for a
cluster analysis, which requires continuous, quantitative data. 8 We considered other
factors that could affect states’ use of Medicaid managed care, such as states’
regulatory environment and staff capacity to oversee Medicaid managed care
programs, but found these to be either difficult to quantify or not readily available,
and we ultimately excluded these factors from our analysis.

Data Reliability

To determine the reliability of the data sources we used, we reviewed related
documentation and, when necessary, interviewed experts most knowledgeable in
the collection and validation of the data. For large electronic data sets that we used,
such as the MSIS APS, we conducted electronic testing for missing data, outliers,
and other apparent errors. For example, we tested whether states known to not use
a certain type of Medicaid managed care, such as MCOs, erroneously reported
enrollment in or capitated payments to such plans. We also compared our results,
when possible, to similar estimates of managed care enrollment available in other
data sources, such as the 2008 Medicaid Managed Care Enrollment Report, which
is derived from the MMCDCS. We also compared our estimates to other publicly
available estimates, such as those compiled by the Kaiser Commission on Medicaid
and the Uninsured and the Medicaid and CHIP Payment and Access Commission.
We determined that the data sources we used were sufficiently reliable for the
purposes of our engagement.

Methodology

After compiling data on the indicators we identified, we used cluster analysis to
identify states that were similar across multiple indicators simultaneously. Cluster
analysis is a method of measuring the degree to which groups of objects—in our
case, states—resemble each other on many different characteristics. One can
measure group similarity on a single characteristic simply by using graphs,
descriptive statistics, or manually inspecting the data. However, these methods are
less useful in describing similarities on multiple characteristics at once. The form of
7
 States were inconsistent in how they defined limited benefit plans or how they reported enrollment
and capitated payments. Cluster analysis requires that all observations in the sample data have
complete information, as missing or incorrect values could skew results.
8
  CMS officials advised us that these are data are the only available program management data
collected from states systematically and stored centrally, and that the data are an important source of
high-level information on states’ activities. However, the officials acknowledged that the data are
limited in their robustness and level of detail.



12                                                          GAO-12-872R Medicaid Managed Care
Enclosure I

cluster analysis we used creates an index from multiple indicators of interest, and
then uses the index to form a sequence of clusters that range from most similar, with
each state as its own cluster, to least similar, with all states in a single cluster. We
did not have strong prior hypotheses about the multivariate distribution of the data,
or where potential clusters may be located. As a result, the nonparametric and
exploratory nature of our clustering method was appropriate to identify many clusters
at varying degrees of similarity.

Because this method uses mathematical clustering rules and measures of similarity,
it identifies potential clusters in a more objective, systematic, and replicable way
than methods that require more human judgment. Cluster analysis requires us to
decide (1) how to measure similarity across multiple variables, and (2) how to
identify clusters of states that are similar to each other. The diversity of our indicator
scales makes the measurement of similarity somewhat difficult. As table 2 shows,
several of our indicators are scaled as proportions, but others are small counts or
broader concepts measured on arbitrary scales. The maximum values across all
indicators range from 2.79 to 10,000.




13                                                  GAO-12-872R Medicaid Managed Care
Enclosure I

Table 2: Descriptive Statistics for Indicators Used in Cluster Analysis

Indicator                  Mean     Minimum     25th Percentile    Median       75th Percentile   Maximum
Population-based Characteristics
Percent of Medicaid
beneficiaries enrolled
in managed care
organizations (MCO)         36.2            0                 0       46.2                65.1        92.9
Percent of Medicaid
beneficiaries with
complex health care
needs enrolled in
MCOs                        16.9            0                 0           7.1             31.2        98.0
Percent of Medicaid
beneficiaries enrolled
in primary care case
management (PCCM)
programs                    20.3            0                 0           6.9             40.8        86.1
Percent of Medicaid
beneficiaries with
complex health care
needs enrolled in
PCCM programs               15.0            0                 0           5.0             22.1        93.9
Medicaid Expansion
Index                        100           25              90.2           101            117.1       153.9
State Market and Other Characteristics
Concentration of low-
income individuals in
urban areas                 74.5         45.2              62.8       75.9                86.4        100
Health maintenance
organization (HMO)
penetration rate            17.2          0.1                7.7      16.6                24.2        54.1
Commercial HMO
market competition
index                      3,889        1,293             2,426      3,414               4,758      10,000
Number of MCOs per
100,000 Medicaid
beneficiaries               0.53            0                 0       0.46                0.76        2.79
Primary Care Capacity
Index                        100         55.5              75.7       85.2               115.7       244.4
Allowable primary care
providers (PCP): MCO         6.0            0                 0           7.5               9.2        13
Allowable PCPs:
PCCM                         5.2            0                 0            6                 9         14
Source: GAO.



To make the indicators comparable, we standardized their scales in two different
ways. First, we normalized each indicator so that its mean and standard deviation
equaled 0 and 1, respectively. Second, we rescaled the indicators so that they
ranged on the unit interval. Because both transformations are affine (or order-
preserving), they eliminated scale differences, while preserving the relative position
of the states in each univariate distribution. This, in turn, allowed us to compare the
relative location of each state in multiple dimensions. After rescaling the data, we
calculated the Euclidean distance between each of the (N*(N-1))/2 = 1,275 possible


14                                                            GAO-12-872R Medicaid Managed Care
Enclosure I

pairs of states and arranged the results in a symmetric matrix. This served as our
final measure of multivariate distance. Using Euclidean distance, the normal
standardization assumes that each variable has equal weight on the overall
distance, but the unit interval standardization relaxes this assumption.

We used a hierarchical, agglomerative algorithm to identify clusters from the
distance matrix. We first assumed that each state was its own cluster, and then
combined states into larger clusters using the “complete linkage” method. This
clustering rule avoided giving too much weight to the moderate number of outliers in
the data, particularly states that enrolled zero participants in Medicaid managed
care, unlike single or, to a lesser extent, average or median linkage methods. Our
method produced a sequence of cluster options that varied from most to least
homogenous within clusters. Working from these many clustering options, we used
our knowledge of Medicaid managed care, and PPACA to create a final set of
clusters that meaningfully described variation across states.

Specifically, our clustering algorithm had the following form. For any multivariate
distance h ≥ 0, let i = 1, 2, … , N index states and j = 1, 2, … , N h index clusters
(subsets of states). Let X ij denote the vector of observations for state i in cluster j.
The complete linkage algorithm initializes with h0 = 0 and i = j. For each j and h > h0,
complete linkage adds state i to cluster j if max(d(X i, X j)) ≤ h for all i in j, where d(∙) is
the Euclidean distance function and the maximum is taken over i. This step of the
algorithm repeats until N h = 1. We then chose from the sequence of clusters formed
along values of h.

Results

Figure 2 presents the results of our cluster analysis. This version of the analysis
used the normalized standardization and all variables in table 2. The figure plots the
distance between each pair of states as a matrix of colors, or a “heat map,” in which
states that are more similar are shaded in darker red and those that are dissimilar
are shaded in yellow. A “dendrogram” above the heat map represents the clustering
process. Each line at the bottom of the dendrogram denotes a single state. The
convergence of lines represents the combination of states into larger clusters at
decreasing levels of similarity, which is measured on the vertical axis. Because
distance matrices are symmetric, the plot is reflected above and below the diagonal.
The results show four broad clusters, or groups, of states, based predominantly on
the proportions of Medicaid beneficiaries enrolled in MCOs or PCCM programs,
relative to other states. Within these large clusters, subclusters form around the
degree of potential Medicaid expansion, the structure of the health care market, and
the concentration of low-income individuals that lived in urban areas.




15                                                      GAO-12-872R Medicaid Managed Care
Enclosure I

Figure 2: Distance Matrix and Results of Cluster Analysis




Because rescaling on the unit interval did not meaningfully change the four broad
clusters we identified, we concluded that our choice of scale was not critical, given
the moderate stability of the results and our focus on the four broad clusters. Our
four broad clusters persisted when removing various indicators from the multivariate
measure of distance. As with rescaling the data, some subclusters included different
states, and some states might have been reclassified as outliers. Ultimately, we
decided to include all of the variables in table 2 in order to allow each variable to
contribute to the final results.




16                                                          GAO-12-872R Medicaid Managed Care
Enclosure II

                     Summary of States’ Use of Medicaid Managed
                         Care by Groups of Similar States

This enclosure highlights the indicators of states’ use of Medicaid managed care that
were generally shared among states in each of the four distinct clusters, or groups,
of states we identified. The descriptions provided in this enclosure generally focus
on the indicators that appeared to have had significant influence on how states
grouped. For labeling purposes, we typically describe the four groups on the basis of
states’ enrollment of Medicaid beneficiaries in managed care organizations (MCO)
and primary care case management (PCCM) programs, which was generally the
predominant similarity among states within each group.

Summary of Group 1: PCCM Predominant States

Group 1 was the largest group of states we identified and included 18 states:
Alabama, Alaska, Arkansas, Idaho, Illinois, Iowa, Louisiana, Maine, Mississippi,
Montana, New Hampshire, North Carolina, North Dakota, Oklahoma, South Dakota,
Utah, Vermont, and Wyoming. These states enrolled, on average, the highest
percentage of beneficiaries in PCCM programs, and typically did not enroll any
Medicaid beneficiaries in MCOs. In addition, these states generally had a low
concentration of low-income individuals that lived in urban areas, and lower
managed care penetration in their overall health insurance markets compared to
states in the other groups we identified. States in Group 1 generally shared the
following characteristics:

High PCCM and No or Low MCO Enrollment

•    PCCM Enrollment: Fourteen of 18 states enrolled Medicaid beneficiaries,
     including those with complex health care needs, in PCCM programs. On
     average, states in Group 1 enrolled 45 percent of all Medicaid beneficiaries in
     PCCM programs, which was higher than the national average of 20 percent, and
     highest among the four groups we identified. (See fig. 3.) In addition, these states
     typically extended enrollment to Medicaid beneficiaries with complex health care
     needs, though to a lesser degree than their general Medicaid population. In
     2008, most of these states enrolled between 20 to 50 percent of Medicaid
     beneficiaries with complex health care needs in PCCM programs. 1




1
 Our definition of “Medicaid beneficiaries with complex health care needs” includes beneficiaries who
were aged, blind/disabled, medically needy, or dually eligible for Medicare and Medicaid. We
excluded from our definition dual eligibles who only received Medicare cost-sharing assistance
through Medicaid.



17                                                         GAO-12-872R Medicaid Managed Care
Enclosure II

Figure 3: Percentage of Medicaid Beneficiaries Enrolled in Primary Care Case Management (PCCM)
Programs among Group 1 States, Fiscal Year 2008




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file. Oklahoma operates an enhanced PCCM program, but the state’s enrollment data are excluded from this
figure because we determined that they were unreliable for our purposes. Specifically, the state reports enrollment in its
enhanced PCCM program as enrollment in a limited benefit plan rather than in the PCCM program category in the MSIS.
Therefore, we could not accurately estimate enrollment numbers in the state’s PCCM program. The following states did not
enroll any Medicaid beneficiaries in PCCM programs and were excluded from this figure: Alaska, Mississippi, New Hampshire,
and Wyoming. However, Alaska, Mississippi, New Hampshire, Oklahoma, and Wyoming were included in our calculation of the
Group 1 average percentage of Medicaid beneficiaries enrolled in PCCM programs.


•    MCO Enrollment: Sixteen of the 18 states did not enroll any Medicaid
     beneficiaries in MCOs in 2008. In the 2 states that did enroll Medicaid
     beneficiaries in MCOs—Illinois and Iowa—less than 10 percent of all
     beneficiaries were enrolled in MCOs, and an even lower percentage of
     beneficiaries with complex health care needs were enrolled.

•    No PCCM or MCO Enrollment: Four states—Alaska, Mississippi, New
     Hampshire, and Wyoming—did not enroll any Medicaid beneficiaries in PCCM
     programs or MCOs in 2008.

Low Concentration of Low-Income Individuals in Urban Areas

Sixteen of the 18 states had a lower percentage of low-income individuals that lived
in urban areas than the national average (75 percent). Most commonly, the overall
percentages of low-income individuals that lived in urban areas was between 45 and
75 percent, which was generally a lower percentage than in states in other groups.




18                                                                     GAO-12-872R Medicaid Managed Care
Enclosure II

Low HMO Penetration

The health maintenance organization (HMO) penetration rate, which measured the
percentage of the total population enrolled in an HMO, such as commercial HMOs,
Medicaid MCOs, and Medicare managed care plans, was lower than the national
average (17.2 percent) in 17 of the 18 states, and was typically less than 10 percent.

Variation in Potential Expansion of Medicaid and Managed Care

These states varied with regard to the degree of potential Medicaid expansion, in
that 8 states were above the national average and 10 states were below. (See
fig. 4.) However, 4 of these states—Alaska, Louisiana, Mississippi, and Oklahoma—
had the highest potential expansion index of all states. Ten states reported plans to
expand the use of Medicaid managed care, 6 of which had a potential Medicaid
expansion that was above the national average. 2




2
  In a 50-state survey, the Kaiser Commission on Medicaid and the Uninsured asked states to
describe the future direction of Medicaid managed care in their states, including any plans to expand
its use. For more information, see K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of
Medicaid Managed Care Programs in 2010: Findings from a 50-State Survey (Washington, D.C.:
Kaiser Family Foundation, Kaiser Commission on Medicaid and the Uninsured, September 2011).



19                                                         GAO-12-872R Medicaid Managed Care
Enclosure II

Figure 4: Group 1 States by Potential Medicaid Expansion Index and Medicaid Managed Care Expansion
Plans




Note: The potential Medicaid Expansion Index values are presented on this map for each state in Group 1. GAO analysis of
data from L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing Primary Care for
Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI: 10.1056/NEJMp1011623;
and K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care Programs in 2010: Findings from
a 50-State Survey (Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on Medicaid and the Uninsured,
September 2011).


Summary of Group 2: States That Use Both MCOs and PCCM Programs

Group 2 was the second largest group of states we identified, and included
16 states: Colorado, Connecticut, Florida, Georgia, Indiana, Kansas, Kentucky,
Missouri, Nebraska, Nevada, Pennsylvania, South Carolina, Texas, Virginia,
Washington, and West Virginia. Most notably, these states were generally
characterized by their enrollment of Medicaid beneficiaries, including those with
complex health care needs, in MCOs and PCCM programs. However, these states
enrolled beneficiaries in MCOs or PCCM programs to a lesser extent than the other
groups of states that used such arrangements. (See fig. 5.) Additionally, when
compared to the other groups of states, Group 2 states generally had an average
concentration of low-income individuals that lived in urban areas and average
managed care penetration.


20                                                                      GAO-12-872R Medicaid Managed Care
Enclosure II

MCO and PCCM Enrollment

•    Above Average MCO Enrollment: All 16 states enrolled Medicaid beneficiaries,
     including those with complex health care needs, into MCOs in 2008. On average,
     these states enrolled 46 percent of all Medicaid beneficiaries in such plans,
     which was 10 percentage points higher than the national average (36 percent),
     but notably lower than the average percentage of beneficiaries enrolled in MCOs
     in Group 3 (64 percent) and Group 4 (67 percent). Over half of these states
     enrolled between 40 and 60 percent of all Medicaid beneficiaries in MCOs, but
     notable variation in enrollment existed among the states. Four states—Colorado,
     Kentucky, Nebraska, and South Carolina—enrolled less than a third of all
     beneficiaries; and two states—Georgia and Indiana—enrolled two-thirds or more
     of all beneficiaries. While all these states extended MCO enrollment to Medicaid
     beneficiaries with complex health care needs, they most commonly enrolled less
     than 20 percent of such beneficiaries in MCOs.

•    Moderate PCCM Enrollment: Thirteen states enrolled Medicaid beneficiaries,
     including those with complex health care needs, in PCCM programs in 2008.
     Enrollment in these states varied widely, ranging from less than 10 percent of all
     Medicaid beneficiaries in about half of the states, to 45 percent of beneficiaries in
     Kentucky. While 8 of the 13 states enrolled beneficiaries with complex health
     care needs to a lesser extent than their total population, 5 states—Colorado,
     Georgia, Indiana, Kansas, and Washington—enrolled a considerably higher
     percentage of beneficiaries with complex health care needs than the percentage
     of overall beneficiaries.




21                                                   GAO-12-872R Medicaid Managed Care
Enclosure II

Figure 5: Percentage of Medicaid Beneficiaries Enrolled in Managed Care Organizations (MCO) and
Primary Care Case Management (PCCM) Programs among Group 2 States, Fiscal Year 2008




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file.


Average Concentration of Low-Income Individuals in Urban Areas

On average, the percentage of low-income individuals that lived in urban areas in
this group of states (76 percent) was roughly equal to the national average of
75 percent, and was over 10 percentage points higher than the average of states in
Group 1 (63 percent). It was also about 10 percentage points lower than the average
of states in Group 3 (85 percent) and Group 4 (87 percent).




22                                                                     GAO-12-872R Medicaid Managed Care
Enclosure II

Average HMO Penetration

This group of states had an average HMO penetration rate (16 percent) that was
roughly equal to the national average of 17 percent. When compared to the other
groups, the average HMO penetration rate in these states was higher than those in
Group 1 (7 percent), but lower than those in Group 3 (26 percent) and Group 4
(36 percent).

Variation in Potential Expansion of Medicaid and Managed Care

These states varied with regard to the degree of potential Medicaid expansion, in
that 10 states were higher than the national average and 6 states were lower. (See
fig. 6.) However, three of these states—Georgia, Kentucky and South Carolina—
were among the 10 states with the largest potential Medicaid expansions. Ten states
reported plans to expand the use of Medicaid managed care in the future, seven of
which had a potential Medicaid expansion that was above the national average. 3




3
See K. Gifford et al., A Profile of Medicaid Managed Care Programs in 2010.



23                                                       GAO-12-872R Medicaid Managed Care
Enclosure II

Figure 6: Group 2 States by Potential Medicaid Expansion Index and Medicaid Managed Care Expansion
Plans




Note: The potential Medicaid Expansion Index values are presented on this map for each state in Group 2. GAO analysis of
data from L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing Primary Care for
Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI: 10.1056/NEJMp1011623;
and K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care Programs in 2010: Findings from
a 50-State Survey (Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on Medicaid and the Uninsured,
September 2011).


Summary of Group 3: MCO Predominant States

The 12 states in Group 3—Arizona, California, Delaware, Maryland, Michigan,
Minnesota, New Jersey, New Mexico, New York, Ohio, Rhode Island, and
Wisconsin—were among those that generally enrolled the highest percentage of
Medicaid beneficiaries in MCOs, and typically did not enroll any Medicaid
beneficiaries in PCCM programs. In addition, these states generally had a high
concentration of low-income individuals that lived in urban areas, and high managed
care penetration within the states’ overall health insurance markets when compared
to the other groups of states we identified.




24                                                                      GAO-12-872R Medicaid Managed Care
Enclosure II

High MCO and No or Low PCCM Enrollment

•    MCO Enrollment: All 12 states enrolled Medicaid beneficiaries, including those
     with complex health care needs, in MCOs in 2008. On average, these states
     enrolled 64 percent of all Medicaid beneficiaries in MCOs, which was
     substantially higher than the national average of 36 percent and the enrollment
     averages in Groups 1 and 2. (See fig. 7.) All 12 states also extended MCO
     enrollment to Medicaid beneficiaries with complex health care needs, though to a
     lesser degree than their general Medicaid population. These states typically
     enrolled between one-third and two-thirds of Medicaid beneficiaries with complex
     health care needs in MCOs.

Figure 7: Percentage of Medicaid Beneficiaries Enrolled in Managed Care Organizations among Group 3
States, Fiscal Year 2008




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file.


•    PCCM Enrollment: Ten of the 12 states did not enroll any Medicaid beneficiaries
     in PCCM programs in 2008. The 2 states that did enroll Medicaid beneficiaries in
     PCCM programs at this time—Delaware and New York—enrolled less than
     10 percent of enrollees in PCCM programs, and enrolled an even lower
     percentage of beneficiaries with complex health care needs in such programs.

High Concentration of Low-Income Individuals in Urban Areas

The average percentage of low-income individuals that lived in urban areas in these
states was 85 percent, which was notably higher than the national average
(75 percent) and the averages in both Group 1 (63 percent) and Group 2
(76 percent).




25                                                                     GAO-12-872R Medicaid Managed Care
Enclosure II

High HMO Penetration

The HMO penetration rate was higher than the national average (17 percent) in all
12 states, and these states, on average, had a higher HMO penetration rate
(27 percent) than that of both Group 1 (7 percent) and Group 2 (16 percent).

Variation in Potential Expansion of Medicaid and Managed Care

These states varied with regard to the degree of potential Medicaid expansion, in
that 7 states were above the national average for potential Medicaid expansion and
5 states were below the national average. (See fig. 8.) New Mexico was among the
10 states with the largest potential degree of Medicaid expansion in the United
States. Five states reported plans to expand the use of Medicaid managed care, one
of which, Michigan, had a potential Medicaid expansion that was above the national
average. 4




4
See K. Gifford et al., A Profile of Medicaid Managed Care Programs in 2010.



26                                                       GAO-12-872R Medicaid Managed Care
Enclosure II

Figure 8: Group 3 States by Potential Medicaid Expansion Index and Medicaid Managed Care Expansion
Plans




Note: The potential Medicaid Expansion Index values are presented on this map for each state in Group 3. GAO analysis of
data from L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing Primary Care for
Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI: 10.1056/NEJMp1011623;
and K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care Programs in 2010: Findings from
a 50-State Survey (Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on Medicaid and the Uninsured,
September 2011).


Summary of Group 4: Other States

Group 4 consisted of five states—the District of Columbia, Hawaii, Massachusetts,
Oregon, and Tennessee. 5 When compared to the other three groups we identified,
the degree of similarity across these five states was the weakest, which may be due
to the fact that one of these states was typically an “outlier” on one or more of the
indicators we examined. That is, at least one state had a value on an indicator that
was either markedly higher or lower than the values for the other four states, thus
diminishing the degree of overall similarity across states in this group. Although
these states exhibited similarities on certain key indicators—such as MCO
enrollment or concentration of low-income individuals that lived in urban areas—to

5
 We use “states” to refer to the 50 states and the District of Columbia.


27                                                                      GAO-12-872R Medicaid Managed Care
Enclosure II

states in the other groups, they did not cluster with those states, possibly as result of
the states’ “outlier” values on indicators.

High MCO Enrollment and No or Low PCCM Enrollment

All five states enrolled Medicaid beneficiaries, including those with complex health
care needs, into MCOs, and with the exception of Massachusetts, did not typically
enroll beneficiaries in PCCM programs. 6 Specifically, these states enrolled an
average of 67 percent of all Medicaid beneficiaries in MCOs, which was the highest
average among all groups. (See fig. 9.) At 93 percent, Tennessee had the highest
MCO enrollment of all states, which was 15 percentage points higher than the next
highest state. All five states also enrolled Medicaid beneficiaries with complex health
care needs in MCOs, but generally to a lesser degree than their overall Medicaid
population. 7 Massachusetts differed from the other 4 states in that it enrolled a
smaller percentage of Medicaid beneficiaries in MCOs, and did enroll a sizeable
percentage of beneficiaries in PCCM programs. Specifically, Massachusetts enrolled
30 percent of all Medicaid beneficiaries in MCOs, and 24 percent of beneficiaries in
PCCM programs, which was similar to MCO and PCCM enrollment of states in
Group 2. In contrast, the other four states in this group—the District of Columbia,
Hawaii, Oregon, and Tennessee—were most similar to Group 3 states in terms of
MCO and PCCM enrollment.




6
 Although Oregon did use a PCCM program in 2008, it enrolled 1 percent of its Medicaid beneficiaries
in this program, which was the second smallest percentage of Medicaid beneficiaries enrolled among
states that used PCCM programs.
7
 Although Hawaii enrolled a high percentage of Medicaid beneficiaries in MCOs, which was similar to
some other states in this group, it enrolled a distinctly smaller percentage of beneficiaries with
complex health care needs (8 percent) than those other states enrolled.



28                                                        GAO-12-872R Medicaid Managed Care
Enclosure II

Figure 9: Percentage of Medicaid Beneficiaries Enrolled in Managed Care Organizations (MCO) among
Group 4 States, Fiscal Year 2008




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file. We use “states” to refer to the 50 states and the District of Columbia.


High Concentration of Low-Income Individuals in Urban Areas

On average, 86 percent of low-income individuals in these states lived in urban
areas, which was the highest concentration of all groups we identified. The District of
Columbia had the highest concentration of low-income individuals that lived in urban
areas (100 percent) of all states. With respect to this indicator, states in this group
generally were similar to states in Group 3; however, Tennessee, where the
concentration of low-income individuals that lived in urban areas was 65 percent,
was most similar to states in Group 1.

High HMO Penetration

All 5 states had HMO penetration rates that were higher than the national average
(17 percent), and three of the states—the District of Columbia, Hawaii, and
Massachusetts—had HMO penetration rates that were among the top 5 states
nationally. At 54 percent, Hawaii had the highest HMO penetration of any other
state, which was 12 percentage points higher than the second highest state.
Generally, states in this group exhibited similar HMO penetration rates as states in
Group 3.




29                                                                     GAO-12-872R Medicaid Managed Care
Enclosure II

Variation in Potential Expansion of Medicaid and Managed Care

Like the other three groups we identified, states in Group 4 varied with regard to the
degree of potential Medicaid expansion in that two states were above the national
average for potential Medicaid expansion and three states were below.
Massachusetts, which implemented a statewide health care reform in 2006 that
included an expansion of Medicaid, was expected to have the smallest potential
Medicaid expansion of any state as a result of Medicaid eligibility expansion
requirements under the Patient Protection and Affordable Care Act. 8 Oregon,
however, was among the 10 states with the largest potential degree of Medicaid
expansion, and reported plans to expand the use of Medicaid managed care. 9




8
 Patient Protection and Affordable Care Act of 2010 (PPACA), Pub. L. No. 111-148, 124 Stat. 119 ,
as amended by the Health Care and Education Reconciliation Act of 2010 (HCERA), Pub. L. No. 111-
152, 124 Stat. 1029. For purposes of this report, references to PPACA include the amendments
made by HCERA.
9
See K. Gifford et al., A Profile of Medicaid Managed Care Programs in 2010.



30                                                       GAO-12-872R Medicaid Managed Care
Enclosure III

                             Summary of Indicators for States

This enclosure provides state-specific data on each of the indicators that we
included in our analysis. We describe each indicator and its relevant data source,
and provide general observations of differences among groups of states with respect
to each indicator.

Percentage of Medicaid Beneficiaries Enrolled in Managed Care Organizations
and Primary Care Case Management Programs

The Medicaid managed care enrollment indicators we identified provide information
on the
•    percentage of Medicaid beneficiaries who were enrolled in a managed care
     organization (MCO); 1

•    percentage of Medicaid beneficiaries with complex health care needs enrolled in
     an MCO; 2

•    percentage of Medicaid beneficiaries enrolled in a primary care case
     management (PCCM) program; 3 and

•    percentage of Medicaid beneficiaries with complex health care needs enrolled in
     a PCCM program.

We obtained data from the Centers for Medicare & Medicaid Services’ (CMS)
Medicaid Statistical Information System, which states report to CMS quarterly, and
which are summarized at an individual level annually in CMS’s Annual Person
Summary (APS) file. Using the APS, we identified the beneficiaries who were
enrolled in MCOs or PCCM programs, including beneficiaries that we considered to
have complex health care needs.




1
 We considered someone as enrolled in an MCO if they were both reported as enrolled and had a
capitated payment made on their behalf to an MCO.
2
 For the purposes of this analysis, “Medicaid beneficiaries with complex health care needs” includes
beneficiaries who were aged, blind/disabled, medically needy, or dually eligible for Medicare and
Medicaid. We excluded dual eligibles who only received Medicare cost-sharing assistance through
Medicaid from our definition.
3
 We considered individuals as enrolled in a PCCM program if they were both reported as enrolled
and had a monthly case management fee paid on their behalf to a PCCM provider.



31                                                         GAO-12-872R Medicaid Managed Care
Enclosure III

Percentage of Medicaid Beneficiaries Enrolled in MCOs

In 2008, 35 states enrolled Medicaid beneficiaries into MCOs, but enrollment varied
widely, ranging from about 1 percent in Iowa to 93 percent in Tennessee. Twenty-
eight states, including all states in Group 3, and 4 of 5 states in Group 4, enrolled
more than one-third of beneficiaries in MCOs, of which 16 states enrolled more than
60 percent. (See fig. 10.) The 35 states that used MCOs in 2008 also enrolled
Medicaid beneficiaries with complex health care needs into MCOs, although to a
lesser extent than the overall Medicaid population in all states except Colorado and
Tennessee.

Figure 10: Percentage of Medicaid Beneficiaries Enrolled in Managed Care Organizations (MCO) in 35
States with MCOs, Fiscal Year 2008




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file. We use “states” to refer to the 50 states and the District of Columbia.




32                                                                     GAO-12-872R Medicaid Managed Care
Enclosure III

Percentage of Medicaid Beneficiaries Enrolled in PCCM Programs

Thirty-one states enrolled Medicaid beneficiaries into PCCM programs, although
enrollment was relatively low except in the Group 1 states, which typically enrolled
more than 50 percent of Medicaid beneficiaries in such programs. 4 (See fig. 11.) All
31 states also enrolled beneficiaries with complex health care needs in PCCM
programs, although generally to a lesser extent than the overall Medicaid population.

Figure 11: Percentage of Medicaid Beneficiaries in Primary Care Case Management (PCCM) Programs in
30 States with PCCM Programs, Fiscal Year 2008




Note: GAO analysis of the Center for Medicare & Medicaid Services’ 2008 Medicaid Statistical Information System Annual
Person Summary file.




4
 Oklahoma operates an enhanced PCCM program, but the state’s enrollment data are excluded from
the data presented here because we determined that they were unreliable for our purposes.
Specifically, the state reports enrollment in its enhanced PCCM program as enrollment in a limited
benefit plan rather than in the PCCM program category in the MSIS. Therefore, we could not
accurately estimate enrollment in the state’s PCCM program.



33                                                                     GAO-12-872R Medicaid Managed Care
Enclosure III

Concentration of Low-Income Individuals in Urban Areas

This indicator measures the percentage of individuals with incomes below
125 percent of the federal poverty level (FPL) that lived in urban areas from calendar
year 2005 to 2009. In general, states with higher MCO enrollment, such as those in
Groups 3 and 4, had a higher concentration of low-income individuals that lived in
urban areas, while states with greater PCCM enrollment and no MCO enrollment,
particularly those in Group 1, generally had a lower concentration of low-income
individuals that lived in urban areas. States with a comparatively small concentration
of low-income individuals in urban areas may face challenges attracting MCOs that
may have concerns about the availability of adequate provider networks or the
sufficiency of enrollment.

Figure 12: Concentration of Low-Income Individuals in Urban Areas by State, Calendar Years 2005-2009




Note: GAO analysis of the U.S. Census Bureau 2005-2009 American Community Survey. We use “states” to refer to the
50 states and the District of Columbia.




34                                                                   GAO-12-872R Medicaid Managed Care
Enclosure III

Health Maintenance Organization Penetration Rate

A state’s health maintenance organization (HMO) penetration rate is the percentage
of the total population in the state that is enrolled in HMOs, such as commercial
HMOs, Medicaid MCOs, and Medicare managed care plans. The state HMO
penetration rates we used were based on population data from the U.S. Census
Bureau as of July 2010. 5 Typically, less than one-third of a state’s population was
enrolled in such plans; however, a larger percentage of Medicaid beneficiaries were
enrolled in MCOs, on average, than the percentage of the total U.S. population
enrolled in any HMOs. In 2008, an average of 36 percent of Medicaid beneficiaries
nationwide was enrolled in an MCO, whereas about 17 percent of the general
population, on average, was enrolled in an HMO in 2010. States with MCOs that had
the highest MCO enrollment, states in Groups 3 and 4, also typically had high HMO
penetration rates. As expected, HMO penetration rates were lowest among states in
Group 1, which generally did not enroll beneficiaries into MCOs.




5
 The Kaiser Family Foundation reports on HMO penetration rate. See Kaiser Family Foundation,
State HMO Penetration Rate, July 2010 (Washington, D.C.: July 2010), accessed September 30,
2011, http://statehealthfacts.org/comparemaptable.jsp?ind=349&cat=7.



35                                                      GAO-12-872R Medicaid Managed Care
Enclosure III

Figure 13: Health Maintenance Organization (HMO) Penetration Rate by State, July 2010




Note: GAO analysis of Kaiser Family Foundation’s state HMO penetration rate data. We use “states” to refer to the 50 states
and the District of Columbia.


Commercial HMO Market Competition Index

This indicator describes the competitiveness of the commercial HMO market as
measured by the Herfindahl-Hisrchman Index (HHI), which was calculated on the
basis of the market shares of commercial HMOs in each state as of January 2009. 6
In general, states with low HHI values—defined as less than 1,500—are considered
to have the most competitive HMO markets, while states with higher HHI values are
6
 We obtained data on state HHI values from the American Medical Association’s (AMA) “Competition
in Health Insurance: A Comprehensive Study of U.S. Markets” 2011 Update, which reported on
market data as of January 1, 2009. The HHI is the sum of the squared market shares of each firm in
the market. See AMA, Competition in health insurance: A comprehensive study of U.S. Markets—
2011 Update (Chicago, Ill.: AMA, Division of Economic and Health Policy Research, 2011).



36                                                                       GAO-12-872R Medicaid Managed Care
Enclosure III

considered to be less competitive. States with HHI values between 1,500 and 2,500
are considered slightly uncompetitive, and those with HHI values greater than 2,500
are considered to be the least competitive. 7

In general, there was variation among states in all four groups with respect to their
HHI values, but overall, most states were not considered to have competitive
commercial HMO markets, according to this measure. Only two states, New York
and Ohio, which are in Group 3, would be considered to have competitive
commercial HMO markets. However, because states in Groups 3 and 4 have been
able to establish and maintain high enrollment in MCOs, the relationship between
the competiveness of a state’s commercial HMO market and its Medicaid MCO
market is unclear.




7
 The HHI, which can reach a maximum value of 10,000, is a commonly used measure of market
concentration and is one of the measures used by the Department of Justice (DOJ) and the Federal
Trade Commission in assessing the effects of mergers on market competition. In general, the more
concentrated a market is, the less competitive it is considered to be. The thresholds for classifying
varying levels of market competitiveness are based on DOJ guidelines.



37                                                         GAO-12-872R Medicaid Managed Care
Enclosure III

Figure 14: Commercial Health Maintenance Organization Market Competition Index by State, January
2009




Notes: GAO analysis of data from American Medical Association (AMA), Competition in health insurance: A comprehensive
study of U.S. markets—2011 Update (Chicago, Ill.: AMA, Division of Economic and Health Policy Research, 2011). Data for
Alaska, Montana, North Dakota, and Wisconsin are excluded above. Data for Alaska, Montana, and North Dakota were
excluded because those states do not have measureable commercial HMO markets. Data for Wisconsin were excluded
because data were unavailable. We use “states” to refer to the 50 states and the District of Columbia.




38                                                                     GAO-12-872R Medicaid Managed Care
Enclosure III

Number of MCOs per 100,000 Medicaid Beneficiaries

This indicator measures the number of MCOs with which each state contracts per
100,000 Medicaid beneficiaries, on the basis of state-reported enrollment data and
the number of MCO contracts in a state as of October 1, 2010. 8 The number of
MCOs per 100,000 Medicaid beneficiaries ranged from 0.12 in Illinois to 2.79 in
Oregon, and averaged 0.75 in the 36 states with MCOs. However, this indicator has
limitations and needs to be considered within the broader context of states’ Medicaid
programs. For example, while Tennessee has the highest MCO enrollment of all
states, the number of MCOs serving Medicaid beneficiaries in the state is among the
lowest because the state purposefully limits the number of MCOs with which they
contract. Similarly, in some states, such as California, certain MCOs operate in
limited regions of the state and are not available to Medicaid beneficiaries outside of
those areas. Nonetheless, on average, groups of states with more MCOs relative to
the size of their Medicaid population also had high percentages of Medicaid
beneficiaries enrolled in MCOs, such as states in Groups 3 and 4.




8
 See K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care
Programs in 2010: Findings from a 50-State Survey (Washington, D.C.: Kaiser Family Foundation,
Kaiser Commission on Medicaid and the Uninsured, September 2011).



39                                                      GAO-12-872R Medicaid Managed Care
Enclosure III

Figure 15: Number of Managed Care Organizations (MCO) per 100,000 Medicaid Beneficiaries in States
Contracting with MCOs, October 2010




Notes: GAO analysis of data from K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care
Programs in 2010: Findings from a 50-State Survey (Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on
Medicaid and the Uninsured, September 2011). GAO used data on the number of MCOs each state contracts with and the total
number of Medicaid beneficiaries in each state to calculate the number of MCOs per 100,000 Medicaid beneficiaries. We use
“states” to refer to the 50 states and the District of Columbia.




40                                                                    GAO-12-872R Medicaid Managed Care
Enclosure III

Medicaid Expansion Index

This indicator provides an estimate of the degree to which states would need to
expand their Medicaid eligibility to fully implement PPACA’s eligibility requirements. 9
The Medicaid expansion index assumed all states would expand Medicaid eligibility
up to 133 percent of the FPL and was developed prior to the U.S. Supreme Court
ruling that states may choose not to expand coverage under PPACA and forgo only
the federal matching funds associated with such expanded coverage. The Medicaid
Expansion Index sets the average of all states at 100; states with index values
greater than 100 have a higher than average potential expansion and states with a
value lower than 100 have a lower than average potential expansion.

We found variation across the states with respect to the Medicaid expansion index,
although some trends emerged. For example, states with higher than average
potential Medicaid expansion index values were largely concentrated in the
Southern region of the country, while states in the Northeast, Midwest, and Western
regions typically had lower than average potential Medicaid expansion index values.
Similarly, on average, states in Groups 1 and 2 had higher than average potential
Medicaid expansion index values when compared to the other 2 groups of states.
For example, 3 of the 4 states in Group 1 that did not enroll any Medicaid
beneficiaries in MCOs or PCCM programs—Alaska, Mississippi, and New
Hampshire—were among the 10 states with the highest expansion index values.




9
 See L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing
Primary Care for Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6
(2011), DOI: 10.1056/NEJMp1011623. The Medicaid expansion index is derived from projections of
the number of individuals considered potentially eligible for Medicaid in 2014 as a result of PPACA’s
expansion of eligibility up to 133 percent of the federal poverty level in relation to the number of low-
income individuals in the state. National Federation of Independent Business, et al., vs. Sebelius,
Sec. of Health and Human Services, et al., No. 11-393 (U.S. June 28, 2012). The extent to which
states will fully implement PPACA’s eligibility expansion is uncertain at this time.



41                                                           GAO-12-872R Medicaid Managed Care
Enclosure III

Figure 16: States’ Medicaid Expansion Index by Group




Note: GAO analysis of data from L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing
Primary Care for Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI:
10.1056/NEJMp1011623. The Medicaid expansion index was based on measures of insurance status and estimates of the
number of nonelderly adults aged 19 to 64 with incomes below 138 percent of the federal poverty level (FPL) from the 2009
and 2010 Annual Social and Economic Supplements (ASEC) of the Current Population Survey (CPS), and projections of the
number of newly eligible Medicaid enrollees based on the 2007 and 2008 ASEC of the CPS and its Health Insurance Policy
Simulation Model (HIPSM). We use “states” to refer to the 50 states and the District of Columbia.




42                                                                     GAO-12-872R Medicaid Managed Care
Enclosure III

Primary Care Capacity Index

The primary care capacity index is a measure of current primary care capacity in
states based on the number of primary care providers (PCP), such as physicians in
general medicine and nurse practitioners, and the number of uninsured patients
served at federally qualified health centers. 10 The average for the index was set at
100 across the states so that states with an index value lower than 100 are
considered to have a lower than average primary care capacity, and those with
index values greater than 100 are considered to have higher than average primary
care capacity. In general, states varied with respect to their primary care capacity
index, but two-thirds of states had lower than average primary care capacity.
Relative to the other groups, states in Groups 1 and 4 had the greatest variation on
this measure, and Group 3 states had the least. All states face the potential for
enrollment increases, particularly if states fully implement PPACA’s Medicaid
eligibility expansion. The increased enrollment will likely increase demand for
primary care, which could be a challenge in states with low primary care capacity.




10
  See L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing
Primary Care for Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6
(2011), DOI: 10.1056/NEJMp1011623. To develop this index, researchers from George Washington
University used data from a variety of sources on the numbers of different types of primary care
providers in 2008 or 2009, and data on the number of patients served at federally qualified health
centers from the Health Resources and Services Administration’s Uniform Data System. Data on the
number of nonfederal physicians in December 2008 are based on estimates from the American
Medical Association; data on the number of nurse practitioners in 2009 are based on the Pearson
Report; and data on the number of physician assistants in clinical practice in December 2008 are
from estimates by the American Academy of Physician Assistants. Data from the Uniform Data
System are for 2009.



43                                                       GAO-12-872R Medicaid Managed Care
Enclosure III

Figure 17: States with an Above Average or Below Average Primary Care Capacity Index by Group




Note: GAO analysis of data from L. Ku, K. Jones, P. Shin, B. Bruen, and K. Hayes, “The State’s Next Challenge—Securing
Primary Care for Expanded Medicaid Populations,” New England Journal of Medicine, vol. 364, no. 6 (2011), DOI:
10.1056/NEJMp1011623. The Primary Care Capacity Index is based on the following data sources: estimates of nonfederal
physicians from the American Medical Association; nurse practitioners in 2009 based on the Pearson Report; projected
number of physician assistants in December 2008 from the American Academy of Physician Assistants; number of patients
served in federally qualified health centers in 2009 from the Uniform Data System of the Bureau of Primary Health Care, Health
Resources and Services Administration. We use “states” to refer to the 50 states and the District of Columbia.




44                                                                       GAO-12-872R Medicaid Managed Care
Enclosure III

Allowable PCP Types

The indicators for allowable PCP types show information on the types of providers
that states permitted MCOs to identify as PCPs or permitted as PCPs in their PCCM
programs as of June 2009. Providers that were considered PCPs included general
practitioners, family practitioners, internists, obstetricians/gynecologists, federally
qualified health centers, rural health clinics, nurse practitioners, nurse midwives,
Indian Health Service providers, physician assistants, psychiatrists, and
psychologists. Within each of the four groups of states we identified, the average
number of different types of providers allowed to participate as PCPs in MCOs or
PCCM programs varied widely. 11 Some states have a greater number of allowable
PCP types, which could be a consideration when assessing a state’s capacity to
provide primary care services to additional Medicaid beneficiaries. 12




11
  We used state-reported data from the CMS Medicaid Managed Care Data Collection System to
calculate the average number of PCP types allowed in a state’s MCO programs. For example,
California reported having 11 MCO programs for which the number of allowable PCPs types ranged
from 6 to 12. We calculated a similar average of the number of allowable PCP types across states’
PCCM programs.
12
  The Primary Care Capacity Index described previously in this report measures availability of certain
providers including physicians, nurse practitioners, physician assistants, and the number of patients
seen at federally qualified health centers; it does not include certain other providers, such as
psychiatrists, psychologists, and social workers that some states report to CMS as being allowable
primary care providers.



45                                                         GAO-12-872R Medicaid Managed Care
Enclosure III

Figure 18: Average Number of Allowable Primary Care Provider (PCP) Types among States with
Managed Care Organizations (MCO) and Primary Care Case Management (PCCM) Programs, June 2009




Notes: GAO analysis of data from the Center for Medicare & Medicaid Services’ 2009 Medicaid Managed Care Data Collection
System and K. Gifford, V. K. Smith, D. Snipes, and J. Paradise, A Profile of Medicaid Managed Care Programs in 2010:
Findings from a 50-State Survey (Washington, D.C.: Kaiser Family Foundation, Kaiser Commission on Medicaid and the
Uninsured, September 2011). Providers that were considered allowable PCPs included general practitioners, family
practitioners, internists, obstetricians/gynecologists, federally qualified health centers, rural health clinics, nurse practitioners,
nurse midwives, Indian Health Service providers, physician assistants, psychiatrists, and psychologists. We use “states” to
refer to the 50 states and the District of Columbia.




46                                                                            GAO-12-872R Medicaid Managed Care
Enclosure IV

                   GAO Contact and Staff Acknowledgments

GAO Contact

Carolyn Yocom, (202) 512-7114 or yocomc@gao.gov

Staff Acknowledgments

In addition to the contact named above, key contributors to this report were Susan
Anthony, Assistant Director; Emily Beller; Julianne Flowers; Joanne Jee; Drew Long;
Katherine Mack; and Jeff Tessin.




(290913)



47                                              GAO-12-872R Medicaid Managed Care
This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.
                      The Government Accountability Office, the audit, evaluation, and
GAO’s Mission         investigative arm of Congress, exists to support Congress in meeting its
                      constitutional responsibilities and to help improve the performance and
                      accountability of the federal government for the American people. GAO
                      examines the use of public funds; evaluates federal programs and
                      policies; and provides analyses, recommendations, and other assistance
                      to help Congress make informed oversight, policy, and funding decisions.
                      GAO’s commitment to good government is reflected in its core values of
                      accountability, integrity, and reliability.

                      The fastest and easiest way to obtain copies of GAO documents at no
Obtaining Copies of   cost is through GAO’s website (www.gao.gov). Each weekday afternoon,
GAO Reports and       GAO posts on its website newly released reports, testimony, and
                      correspondence. To have GAO e-mail you a list of newly posted products,
Testimony             go to www.gao.gov and select “E-mail Updates.”

Order by Phone        The price of each GAO publication reflects GAO’s actual cost of
                      production and distribution and depends on the number of pages in the
                      publication and whether the publication is printed in color or black and
                      white. Pricing and ordering information is posted on GAO’s website,
                      http://www.gao.gov/ordering.htm.
                      Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
                      TDD (202) 512-2537.
                      Orders may be paid for using American Express, Discover Card,
                      MasterCard, Visa, check, or money order. Call for additional information.
                      Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Connect with GAO      Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts.
                      Visit GAO on the web at www.gao.gov.
                      Contact:
To Report Fraud,
Waste, and Abuse in   Website: www.gao.gov/fraudnet/fraudnet.htm
                      E-mail: fraudnet@gao.gov
Federal Programs      Automated answering system: (800) 424-5454 or (202) 512-7470

                      Katherine Siggerud, Managing Director, siggerudk@gao.gov, (202) 512-
Congressional         4400, U.S. Government Accountability Office, 441 G Street NW, Room
Relations             7125, Washington, DC 20548

                      Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
Public Affairs        U.S. Government Accountability Office, 441 G Street NW, Room 7149
                      Washington, DC 20548




                        Please Print on Recycled Paper.