Performance audit, Arizona Health Care Cost Containment System Medicaid eligibility determination

              A REPORT
TO THE
ARIZONA LEGISLATURE
Debra K. Davenport
Auditor General
Performance Audit
Arizona Health Care
Cost Containment
System—Medicaid
Eligibility Determination
Performance Audit Division
June • 2012
REPORT NO. 12-02
The Auditor General is appointed by the Joint Legislative Audit Committee, a bipartisan committee composed of five senators and five
representatives. Her mission is to provide independent and impartial information and specific recommendations to improve the opera-tions
of state and local government entities. To this end, she provides financial audits and accounting services to the State and political
subdivisions, investigates possible misuse of public monies, and conducts performance audits of school districts, state agencies, and
the programs they administer.
The Joint Legislative Audit Committee
Audit Staff
Copies of the Auditor General’s reports are free.
You may request them by contacting us at:
Office of the Auditor General
2910 N. 44th Street, Suite 410 • Phoenix, AZ 85018 • (602) 553-0333
Additionally, many of our reports can be found in electronic format at:
www.azauditor.gov
Dale Chapman, Director
Dot Reinhard, Manager and Contact Person
Steven Meyeroff, Team Leader Lois Sayrs, Senior Methodologist
Mike Devine Diane Schaller, Technology Information
Angela Fanzo Specialist Manager
Breck Johnson
Representative Carl Seel, Chair
Representative Tom Chabin
Representative Justin Olson
Representative David Stevens
Representative Anna Tovar
Representative Andy Tobin (ex officio)
Senator Rick Murphy, Vice Chair
Senator Andy Biggs
Senator Rich Crandall
Senator Linda Lopez
Senator David Lujan
Senator Steve Pierce (ex officio)
2910 NORTH 44th STREET • SUITE 410 • PHOENIX, ARIZONA 85018 • (602) 553-0333 • FAX (602) 553-0051
MELANIE M. CHESNEY
DEPUTY AUDITOR GENERAL
DEBRA K. DAVENPORT, CPA
AUDITOR GENERAL
STATE OF ARIZONA
OFFICE OF THE
AUDITOR GENERAL
June 6, 2012
Members of the Arizona Legislature
The Honorable Janice K. Brewer, Governor
Mr. Tom Betlach, Director
Arizona Health Care Cost Containment System
Mr. Clarence Carter, Director
Department of Economic Security
Transmitted herewith is a report of the Auditor General, A Performance Audit of the Arizona
Health Care Cost Containment System (AHCCCS)—Medicaid Eligibility Determination. This
report is in response to an October 26, 2010, resolution of the Joint Legislative Audit
Committee. The performance audit was conducted as part of the sunset review process
prescribed in Arizona Revised Statutes §41-2951 et seq. I am also transmitting within this
report a copy of the Report Highlights for this audit to provide a quick summary for your
convenience.
As outlined in its response, AHCCCS agrees with all of the findings and plans to
implement all of the recommendations. The report also includes a response from the
Department of Economic Security.
My staff and I will be pleased to discuss or clarify items in the report.
This report will be released to the public on June 7, 2012.
Sincerely,
Debbie Davenport
Auditor General
Attachment
Majority of AHCCCS program operates
under managed care model—Medicaid
is a federal healthcare program for certain
low-income individuals and families that is
jointly funded by federal and state
governments. AHCCCS is Arizona’s state
program that provides these benefits to
eligible persons primarily through a
managed care system. Under this system,
AHCCCS contracts with health plans,
which coordinate and pay for the medical
services AHCCCS members receive from
healthcare providers. To cover the costs of
coordinating and paying for members’
healthcare, the contracted health plans
receive monthly capitation payments for
each enrolled member.
AHCCCS and DES determine
eligibility for the Medicaid
program—To receive Medicaid
services, all applicants must
satisfy various Medicaid
eligibility requirements and be
approved through an eligibility
determination process.
2012
June • Report No. 12-02
Arizona Health Care
Cost Containment
System—Medicaid
Eligibility Determination
Our Conclusion
The Arizona Health Care
Cost Containment System
(AHCCCS), which operates
the State’s Medicaid
program, shares
responsibility for
determining applicant
eligibility with the
Department of Economic
Security (DES). AHCCCS
and DES accurately
determined eligibility for
almost all Medicaid
applicants. However, we
calculated that 5.92
percent of the eligibility
determinations are at risk
for processing errors, and
1.11 percent of eligibility
determinations are at risk
for being incorrect.
Therefore, we estimated
that AHCCCS is paying
between approximately
$3.5 and $4.8 million in
monthly capitation
payments for enrolled but
ineligible members.
AHCCCS should
implement a corrective
action plan to address the
errors that occur in
calculating and verifying
income and resources.
AHCCCS should also
increase its electronic
verification of applicants’
citizenship.
REPORT
HIGHLIGHTS
PERFORMANCE AUDIT
Medicaid applicants must be approved through eligibility
determination process
Eligibility requirements are established by
federal regulations and state law and
require documentation of U.S. citizenship
or qualified alien status, state residency,
and income at or below the prescribed
federal poverty level threshold. In Arizona,
AHCCCS and DES share responsibility for
determining eligibility for Medicaid
applicants. As of July 1, 2011, DES
performed approximately 82 percent of
the eligibility determinations and AHCCCS
performed about 11 percent of them.
About 7 percent of AHCCCS members
are automatically eligible for Medicaid
services, such as children born to women
who are on Medicaid.
1 The 5.92 percent is a weighted error rate we calculated. See the report’s Appendix B, pages b-iii through b-iv.
Number of Eligibility Determinations
and Percentage Approved as Eligible
December 1, 2011 through December 31, 2011
AHCCCS should take some additional actions to strengthen
eligibility determination process
5.92 percent of eligibility determination
cases are at risk for processing
errors—Although AHCCCS and DES
appropriately determined the eligibility of
approximately 94 percent of applicants,
5.92 percent of eligibility determination
cases are at risk for processing errors
based on our review of a representative
sample of 279 eligibility determinations.1
We found that 16 of the 279 eligibility
determinations had processing errors,
including 9 that did not have verification or
documentation of income. For example, in
one determination, the caseworker relied
on the applicant’s statement of income
instead of documentation from the
employer. Subsequent documentation
showed that the applicant understated
monthly income by approximately $380. In
another case, the applicant’s monthly
Agency
Determinations
Performed
Percentage
Approved
DES 261,854 55.7%
AHCCCS 13,815 58.7
Total 275,669 55.8
Arizona Health Care
Cost Containment
System—Medicaid
Eligibility Determination
REPORT
HIGHLIGHTS
PERFORMANCE AUDIT
June 2012 • Report No. 12-02
A copy of the full report is available at:
www.azauditor.gov
Contact person:
Dot Reinhard (602) 553-0333
income was understated by about $650.
For 7 of the 16 determinations with errors, case-workers
miscalculated the amount of income or
resources. Eligibility is based on monthly income,
and errors may occur when the caseworker
converts weekly or biweekly income into monthly
income. For example, in one determination, the
caseworker incorrectly entered a monthly income in
the computer system that was approximately $1,065
less than should have been entered.
Processing errors create potential for incorrect
eligibility determinations—For 13 of the 16
determinations with errors, the processing errors did
not result in an incorrect Medicaid eligibility
determination. However, in three cases, the
processing errors were of sufficient magnitude to
result in incorrect eligibility determinations. In these
instances, the caseworkers did not properly verify or
correctly calculate the income amount used to
make the eligibility determination. As a result,
AHCCCS inappropriately paid $2,359 in capitation
payments for these three members over a 7-month
period.
Although the number of incorrect eligibility determi-nations
was a small part of the sample cases, they
become more important when projected over all
Medicaid eligibility determinations AHCCCS and
DES performed. Based on the incorrect eligibility
determinations identified in the sample, we calculat-ed
an incorrect eligibility determination rate of 1.11
percent for the acute care and long-term care
programs tested. This means that of the approxi-mately
$414 million in monthly capitation payments
that AHCCCS makes for its members in the
programs tested, we estimated that AHCCCS is
paying its health plans approximately $3.5 to $4.8
million monthly for enrolled but ineligible members.
Most error types consistent with federal review—
Every 3 years, the U.S. Department of Health and
Human Services, Centers for Medicare and
Medicaid Services, conducts a Payment Error Rate
Measurement (PERM) review to evaluate the
accuracy of Medicaid eligibility determinations. The
types of errors we identified were similar to the
eligibility determination errors identified by the
federal fiscal year 2008 PERM review. This PERM
review found that Arizona’s error rate was 2.2
percent, which was below the national eligibility
error rate of 6.7 percent for the 17 states reviewed
that year. AHCCCS received another PERM review
in federal fiscal year 2011, and the results are
expected in November 2012.
Monies cannot be recovered, but AHCCCS
should implement corrective action plan—
AHCCCS should develop a corrective action plan
that will help it and DES to correct the kinds of
processing errors this audit identified as well as to
minimize their frequency going forward. Minimizing
errors is important because AHCCCS cannot
recover the approximately $3.5 to $4.8 million in
monthly capitation payments made to health plans
and providers for ineligible members after AHCCCS
and DES officially determined that those members
were eligible for Medicaid services unless the
information used to make the determinations is
proven to be fraudulent. The corrective action plan
should include additional caseworker training in
areas that are error-prone, and AHCCCS should
also assess whether its income and resource
policies need clarification.
AHCCCS and DES should make greater use of
electronic matching to verify citizenship
requirements—Although caseworkers consistently
verified social security numbers using electronic
matching, they did not consistently use electronic
matching to verify income or citizenship. DES has
two electronic income verification systems it can
use to make such verifications and in June 2011, it
enhanced its use of one of these systems for
income verification. However, both AHCCCS and
DES should make greater use of electronic
matching to verify citizenship. In addition, the 2010
Federal Patient Protection and Affordable Care Act
requires states to expand their use of electronic
verification by 2014, including interfaces with the
Social Security Administration for citizenship
verification and the Department of Homeland
Security for immigration status.
Recommendations:
AHCCCS should:
• Develop a corrective action plan to minimize
eligibility determination errors.
• Ensure that it and DES make greater use of
electronic means to verify citizenship.
TABLE OF CONTENTS
continued
page i
Office of the Auditor General
Introduction 1
Finding 1: AHCCCS should take some additional actions to
strengthen the eligibility determination process 9
5.92 percent of individual eligibility determinations are at risk
for processing errors 9
Processing errors create potential for incorrect eligibility determinations 12
Most error types consistent with federal review 13
AHCCCS should implement plan to reduce errors made during
eligibility determinations 13
AHCCCS should increase its use of electronic matching to verify
citizenship requirements 14
Electronic verification will be required by the Affordable Care Act
and is being increasingly adopted by other states to verify citizenship 16
Recommendations 17
Appendix A: Methodology a-i
Appendix B: Sampling, error rate estimate, and
projection methodology b-i
Process for selecting the sample b-i
Steps for estimating processing errors b-iii
Process for calculating an incorrect eligibility determination rate b-iv
TABLE OF CONTENTS
concluded
page ii
State of Arizona
Steps for estimating incorrect capitation payments b-v
Sample demographics b-vii
Statistical glossary b-viii
Agency Response
Tables
1 AHCCCS Enrollment by Program
At July 1, 2009, 2010, 2011, and May 1, 2012 3
2 Schedule of Revenues, Expenditures, and Changes in Fund Balance
Fiscal Years 2010 through 2012
(In Thousands)
(Unaudited) 4
3 Number of Eligibility Determinations and Percentage Approved as Eligible
December 1, 2011 through December 31, 2011 5
4 Division of Member Services Eligibility Determination Expenditures by
Office and Transfers to DES
Fiscal Year 2011
(Unaudited) 8
5 Income Verification Sources Used by AHCCCS and DES to Confirm
Reported Income
April 2011 10
6 Citizen and Qualified Alien Status Verification Sources Used by
AHCCCS and DES
April 2011 16
Medicaid applicants must be approved
through an eligibility determination process
Majority of AHCCCS program operates under
managed care model
AHCCCS was established to administer Arizona’s Medicaid program, which
provides healthcare for certain low-income individuals and families living in
Arizona. Medicaid is a federal healthcare program for low-income individuals
and families that is jointly funded by the federal and state governments.
AHCCCS was implemented in October 1982 as the nation’s first state-wide
Medicaid program designed to provide medical services to eligible persons
primarily through a managed care system. Under a managed care system,
AHCCCS contracts with entities, known as health plans, which coordinate and
pay for the medical services AHCCCS members receive from registered
AHCCCS healthcare providers, such as physicians and hospitals. To cover the
costs of coordinating and paying for members’ healthcare, the contracted
health plans receive monthly capitation payments (see textbox).
page 1
The Office of the Auditor
General has conducted a
performance audit of the
Arizona Health Care Cost
Containment System
(AHCCCS) pursuant to an
October 26, 2010,
resolution of the Joint
Legislative Audit
Committee. This audit is
the second in a series of
audits conducted as part of
the sunset review process
prescribed in Arizona
Revised Statutes (A.R.S.)
§41-2951 et seq and
focuses on whether the
State’s Medicaid eligibility
determinations meet
federal and state
requirements. The first
audit found that AHCCCS
has processes in place that
help it comply with state
and federal requirements
for coordinating the
payment of healthcare
benefits with other
responsible parties. The
third report will examine
AHCCCS’ processes for
managing Medicaid fraud
cases. A fourth report will
address the statutory
sunset factors.
Office of the Auditor General
Scope and Objectives
INTRODUCTION
Capitation payment—A fixed monthly amount paid in advance to AHCCCS’
contracted health plans for each enrolled member. At least annually, based on information
such as the historical use and cost of medical services provided and inflation data,
capitation payment amounts are determined using mathematical and statistical methods.
Monthly capitation amounts paid to AHCCCS’ contracted health plans can vary by
individual based on factors such as age, gender, geographical service area, and program
(see examples below):
1 See page 2 for explanation of Acute Care and Arizona Long Term Care System programs.
Source: Auditor General staff analysis of AHCCCS information contained in its contracts, actuarial
certifications, and Acute Care and Arizona Long Term Care System rates effective October 1, 2011.
Capitation payment
Average
Arizona Long
Term Care
System
monthly
capitation
rate1
Examples of average Acute Care monthly capitation rates1
Age
<1
Male/Female
Age
1-13
Male/Female
Age
14-44
Female
Age
14-44
Male
Age
45+
Male/Female
$460 $97 $222 $138 $347 $3,000
Approximately 90 percent of AHCCCS’ members are enrolled with its contracted
health plans in managed care. For the remaining members, known as fee-for-service
members, AHCCCS reimburses registered healthcare providers directly.1 According
to the Kaiser Family Foundation, as of October 2010, 47 states and the District of
Columbia used managed care programs to some degree, but only 9 states, including
Arizona, had 80 percent or more of their members enrolled in comprehensive
managed care programs.2,3
AHCCCS members receive a full range of medical services under the following three
primary programs:
• Acute Care—As shown in Table 1 (see page 3), the majority of AHCCCS’
members are enrolled in its Acute Care program. This Medicaid program
provides a wide range of healthcare services, such as inpatient and outpatient
hospital services, physician services, immunizations, and laboratory and x-ray
services to children, pregnant women, and other low-income adults.
• Arizona Long Term Care System (ALTCS)—A small percentage of members
receive services under ALTCS. The ALTCS program provides acute care,
behavioral health, long-term care, and case management services to individuals
who are elderly, physically disabled, or developmentally disabled and meet the
criteria for institutionalization.
• KidsCare—Children under age 19 may receive medical services under
KidsCare, the name given to Arizona’s federal Children’s Health Insurance
Program. Children may qualify for KidsCare if their family’s income exceeds the
limit allowed for Medicaid, but is still below the federally established amount for
this program. Children enrolled in KidsCare receive the same medical services
available under Arizona’s Acute Care program. New enrollment in the KidsCare
Program has been frozen since January 1, 2010, due to lack of funding, and
AHCCCS has established a waiting list of applicants. However, effective May 1,
2012 through January 1, 2014, AHCCCS will be receiving monies from three
hospitals that will allow AHCCCS to provide coverage for 21,700 children in what
is being called KidsCare II.4 This state-wide program will offer the same benefits,
but has a lower income eligibility threshold than the KidsCare program.
1 AHCCCS reimburses providers on a fee-for-service basis for (1) individuals receiving services under the Federal
Emergency Services program, or (2) Native American members who choose to receive services through a tribal fee-for-
service contractor.
2 Comprehensive managed care is defined as inpatient hospital services and any of the following services, or any three
of the following services: (1) outpatient hospital services; (2) rural health clinic services; (3) Federally Qualified Center
services; (4) other laboratory and x-ray services; (5) nursing facility services; (6) early and periodic screening,
diagnostic, and treatment services; (7) family planning services; (8) physicians’ services; and (9) home health services.
3 Kaiser Family Foundation. (2010). Medicaid enrollment in comprehensive managed care as a share of total Medicaid
enrollment, October 2010. Retrieved January 31, 2012, from www.statehealthfacts.org
4 Laws 2011, Ch. 234, §2 allows AHCCCS, subject to U.S. Department of Health and Human Services, Centers for
Medicare and Medicaid Services (CMS) approval, to authorize any political subdivision to provide monies necessary
to qualify for federal matching monies to provide healthcare coverage to persons who would have been eligible
pursuant to A.R.S. §36-2901.01.
page 2
State of Arizona
AHCCCS receives federal monies along with state, county, and other monies, such as
tobacco taxes, to operate Arizona’s Medicaid program. As shown in Table 2 (see page
4), during fiscal year 2012, AHCCCS estimates that its revenues will total more than
$8.4 billion, with approximately $5.66 billion coming from the federal government,
approximately $2.16 billion from the State, about $341 million from the counties, and
$275 million from other sources. AHCCCS’ estimated expenditures for fiscal year 2012
total nearly $8.4 billion, with about $6.4 billion, or 76 percent, going toward capitation
payments. AHCCCS’ estimated revenues and expenditures for fiscal year 2012 are
each approximately $1.2 billion less than fiscal years 2010 and 2011 because some
changes were made to Arizona’s Medicaid program during the 2011 legislative
session. For example, enrollment in Arizona’s Medicaid program for some individuals,
such as childless adults, is no longer being accepted.1 In addition, the federal
matching rate returned to its typical level starting in fiscal year 2012. Specifically, the
American Recovery and Reinvestment Act of 2009 and additional federal legislation
increased the federal matching rate from approximately 66 percent to between 71 and
76 percent from October 1, 2008 through June 30, 2011. This change and the changes
to the Arizona Medicaid program resulted in the fiscal year 2012 estimated federal
government revenues being approximately $1.4 billion lower. However, the State’s
estimated revenue did not show a similar decrease, in part due to the reduction in the
federal matching rate that required the State to contribute more of each dollar spent.
States must determine eligibility for Medicaid services
To receive Medicaid services, all applicants must satisfy various Medicaid eligibility
requirements and be approved for Medicaid through an eligibility determination
process. AHCCCS shares the responsibility for determining applicant eligibility with the
Department of Economic Security (DES). Federal regulations and state laws establish
the various eligibility requirements, including income thresholds, citizenship, and state
1 In December 2011, the Arizona Court of Appeals upheld the State’s decision to stop new enrollment for childless adults,
indicating that it was a political decision that was not subject to judicial resolution. In February 2012, the Arizona Supreme
Court refused to review the Appeals Court’s decision.
page 3
Office of the Auditor General
Table 1: AHCCCS Enrollment by Program
At July 1, 2009, 2010, 2011, and May 1, 2012
Source: Auditor General staff analysis of the AHCCCS July 1, 2009, 2010, and 2011, and May 1, 2012, Population
Highlights reports.
AHCCCS receives federal,
state, county, and other
monies, such as tobacco
taxes, to operate Arizona’s
Medicaid program.
Program
Acute Care
Arizona Long Term Care System 48,673 50,241 51,314 52,253
KidsCare 51,838 30,445 17,649 10,966
Total 1 ,369,637 1 ,286,796
1 ,223,577
2009 2010 2011 2012
1 ,174,598
1 ,275,109
1 ,272,118
1 ,352,804
1 ,300,674
page 4
State of Arizona
1 The table includes all AHCCCS financial activity except the Healthcare Group. The Healthcare Group provides medical coverage
primarily to small, uninsured businesses and is managed as a self-supporting operation.
2 The estimates for fiscal year 2012 revenues and expenditures are significantly less than fiscal years 2010 and 2011 because multiple
changes were made to the Medicaid program and the State’s contribution during the 2011 legislative session that affected fiscal year
2012. See page 3 for additional information.
3 Consists of all monies that originally came from the federal, state, or county governments, including monies passed through other
entities, such as other state agencies.
4 Amounts primarily consist of monies that were authorized for use on AHCCCS expenditures by the Legislature or voters, such as
tobacco litigation monies, gaming revenues, and tobacco tax monies administered by AHCCCS. For example, Proposition 204
(November 2000) authorized the use of tobacco settlement monies to increase the number of people eligible for coverage in AHCCCS.
Similarly, Proposition 202 (November 2002) provides a portion of gaming revenues to be used for a trauma and emergency services
program.
5 Amounts consist of capitated mental health and Children’s Rehabilitation Services expenditures that were passed through to the Arizona
Department of Health Services. Beginning in fiscal year 2012, the Children’s Rehabilitation Services appropriation was moved to
AHCCCS; therefore, AHCCCS no longer passes through these monies to the Department and instead makes payments directly to the
providers.
6 Amounts consist of various other expenditures that were not paid as capitated payments or fee-for-service. For example, reinsurance,
a stop-loss program for partial reimbursement after a deductible is met, is included in this category.
7 Amounts primarily consist of monies transferred to the Arizona Departments of Health Services and Economic Security for monies
appropriated by the Legislature to these agencies. Specifically, the Legislature appropriated over $35 million each year in fiscal years
2010 through 2012 to the Department of Health Services for behavioral health services from the tobacco tax monies AHCCCS
administers. Similarly, approximately $3 million each year was appropriated to the Department of Economic Security in fiscal years 2010
through 2012 from county contributions for administration costs for Proposition 204 (November 2000) implementation.
Source: Auditor General staff analysis of the AHCCCS fiscal year 2010 and 2011 financial statements audited by an independent certified
public accounting firm and AHCCCS-prepared fiscal year 2012 estimates dated January 24, 2012, that are primarily composed of
fiscal year 2012 appropriations.
Table 2: Schedule of Revenues, Expenditures, and Changes in Fund Balance1
Fiscal Years 2010 through 2012
(In Thousands)
(Unaudited)
2010 2011 20122
(Actual) (Actual) (Estimate)
Revenues:
Federal government3 $ 7,229,797 $ 7,077,440 $ 5,663,201
State government3 1,720,054 2,012,179 2,163,412
County government3 247,043 277,663 341,131
Other4 302,363 272,449 275,024
Total revenue 9,499,257 9,639,731 8,442,768
Expenditures and transfers:
Capitated payments—
Acute care 4,181,191 4,163,405 3,150,673
Long-term care 1,940,629 1,957,650 1,959,774
KidsCare 91,795 55,095 36,068
Mental health and Children's Rehabilitation Services5 1,413,917 1,422,241 1,234,025
Fee-for-service—
Acute care 847,605 874,121 759,836
Long-term care 119,705 127,138 134,366
Other6 685,871 805,577 898,318
Administrative 177,092 163,936 180,616
Total expenditures 9,457,805 9,569,163 8,353,676
Transfers to the State General Fund 2,699 1,268 1,244
Net transfers to other state agencies7 39,213 38,184 41,928
Total expenditures and transfers 9,499,717 9,608,615 8,396,848
Net change in fund balance ( 460) 31,116 45,920
Fund balance, beginning of year 685 225 31,341
Fund balance, end of year $ 225 $ 31,341 $ 77,261
page 5
Office of the Auditor General
residency. As the State’s Medicaid agency, AHCCCS is ultimately responsible for
ensuring the accuracy of Arizona’s eligibility determinations and has established
processes to help ensure accuracy, such as caseworker training and quality control
reviews of its own and DES’ eligibility determinations.
AHCCCS and DES determine eligibility for the Medicaid program—
AHCCCS, as the State’s Medicaid agency, is required to determine eligibility for the
Medicaid program. However, as allowed, AHCCCS has entered into an intergovern-mental
agreement with DES to perform Medicaid eligibility determinations on its
behalf. DES performs Medicaid eligibility determinations in conjunction with eligibil-ity
determinations for other federal public assistance programs, such as the
Supplemental Nutrition Assistance Program (formerly called food stamps) and the
Temporary Assistance for Needy Families program. Although DES performed
approximately 82 percent of the Medicaid eligibility determinations as of July 1,
2011, it performs the eligibility determinations for only one program: AHCCCS’
Acute Care program. Typical applicants for this program include families with
dependent children, pregnant women, and some adults who do not have children.
AHCCCS performed about 11 percent of the eligibility determinations as of July 1,
2011. It completes determinations for ALTCS, KidsCare, and the Supplemental
Security Income-Medical Assistance Only population, which is part of the Acute
Care program. Additionally, about 7 percent of AHCCCS members are automatically
eligible for Medicaid services such as children born to women who are on Medicaid
or individuals who have been determined
eligible for other programs, such as, aged,
blind, or disabled individuals in the U.S.
Social Security Administration’s
Supplemental Security Income Cash
program. In most cases, as prescribed in
federal regulation, both AHCCCS and DES
must redetermine the eligibility of Medicaid
recipients at least every 12 months.1
According to AHCCCS and DES information,
as illustrated in Table 3, more than 275,000
combined initial and renewal determinations
were completed in December 2011, with
approximately 56 percent being determined
eligible.
Federal regulations and state laws establish Medicaid eligibility
requirements—AHCCCS and DES must determine each applicant’s eligibility in
accordance with federal and state requirements. These requirements focus on
applicants’ financial status and other conditions, such as legally residing in the
United States. These federal eligibility criteria are incorporated in AHCCCS’ feder-
1 In October 2011, CMS approved Arizona’s Medicaid program for another 5-year period. However, CMS did not approve
AHCCCS to conduct redeterminations every 6 months instead of every 12 months as requested by the State.
Table 3: Number of Eligibility Determinations and
Percentage Approved as Eligible
December 1, 2011 through December 31, 2011
1 This number does not include AHCCCS members who are automatically eligible
for Medicaid services, such as individuals in the U.S. Social Security
Administration’s Supplemental Security Income Cash program.
Source: Auditor General staff analysis of initial and renewal AHCCCS- and DES-prepared
application reports.
Agency
Determinations
Performed
Percentage
Approved
DES 261,854 55.7%
AHCCCS 13,815 58.7
Total 275,6691 55.8
page 6
State of Arizona
ally approved State Plan (see textbox). Specific
requirements include:
• United States citizen or qualified alien—
Applicants must reside in the United States
as either a citizen or a qualified alien, who is
a person admitted to the U.S. legally in a
specific classification.1 Documents that
prove citizenship or qualified alien status
include a U.S. passport, naturalization
certificate, permanent resident card, or U.S.
public birth record. AHCCCS and DES can
verify that an applicant is a U.S. citizen or
qualified alien through electronic means, such as an electronic match with
Arizona Vital Records, or by obtaining original documents from applicants
such as birth certificates that can be copied and maintained in its files.
• Verified social security number—Applicants must have a verified social
security number with the U.S. Social Security Administration. AHCCCS and
DES electronically verify this by matching the applicant’s social security
number, name, and date of birth provided by the applicant with data from the
U.S. Social Security Administration.
• Arizona resident—Applicants must be residents of Arizona. AHCCCS and
DES verify residency with documents including rent or mortgage receipts, an
Arizona driver’s license, or utility bills or receipts, but consistent with federal
regulation, will also accept the applicant’s residency declaration on the
application unless there is evidence to the contrary.
• Income limits—Applicants’ or their family’s income must be at or below a
specific percentage of the federally established poverty guidelines, commonly
referred to as the federal poverty level (FPL). The FPL threshold varies by age
and other factors, such as whether the applicant is pregnant. For example, as
of April 2011, the threshold for families with children in the Acute Care program
was 100 percent of the FPL, which means that the monthly income limit for a
family of four is $1,863. In addition, with approval from the federal government,
states can expand their Medicaid programs to cover people whose coverage
is not required by the federal government. For example, in the November 2000
general election, voters expanded eligibility for Arizona’s Medicaid program.
This expansion, outlined in A.R.S. §36-2901.01, is commonly referred to as
Proposition 204. Proposition 204 expanded Arizona’s Medicaid coverage to
all eligible individuals, including childless adults, with incomes at or below 100
percent of the FPL. However, due to state budget difficulties, new enrollments
1 According to federal law, aliens who are legally in the U.S. but who are not qualified aliens and aliens who are not legally
in the U.S. may be determined eligible to receive emergency services, but do not qualify for full AHCCCS coverage.
State Plan—The State Plan is a
comprehensive written statement
submitted by AHCCCS and
approved by the U.S. Department
of Health and Human Services,
Centers for Medicare and
Medicaid Services describing the
nature and scope of Arizona’s
Medicaid program.
Source: OMB Circular A-133, Compliance
Supplement March 2011.
page 7
Office of the Auditor General
in this program were stopped on July 8, 2011.1 AHCCCS and DES staff verify
applicants’ income through electronic or other means, such as pay stubs; and
they must also document their verification.
• Signed application—Applicants are required to sign, under the penalty of
perjury, an application. Applicants can apply for Medicaid benefits through a
variety of means, such as online, in person at an AHCCCS or DES office, or by
mail.
After an application is received, the caseworker verifies the information provided and
either approves or denies the applicant for Medicaid benefits.2 AHCCCS and DES
must maintain a case record of the facts used to support its eligibility determinations.
AHCCCS has established processes to help ensure eligibility deter-minations
meet requirements—As the State’s Medicaid agency, AHCCCS
is ultimately responsible for ensuring the accuracy of Arizona’s eligibility determina-tions
and has established processes to help ensure accuracy. For example,
AHCCCS and DES have established extensive policies and procedures that help
guide staff through the eligibility determination process. In addition, both agencies
provide training to staff on the process. For example, classes cover topics such as
income, citizenship, and residency. AHCCCS is also responsible for conducting two
federally required quality control reviews, one on a monthly basis and the other every
three years, to determine the accuracy of its and DES’ eligibility determinations.
Following these reviews, AHCCCS and DES must identify the reason errors occurred
and take steps to prevent similar errors from occurring in the future.
Staffing and expenditures
AHCCCS and DES have allocated 2,841 full-time equivalent (FTE) positions to the
eligibility determination function. Specifically, as of December 1, 2011, AHCCCS
reported that it allocated 533 FTE positions to eligibility determination. Of these
positions, 26 were vacant for a vacancy rate of just under 5 percent. Seventy-nine
percent of the filled positions were for caseworkers and the rest included support staff
such as quality control, training, and policy staff. In addition, as of September 16, 2011,
DES reported that it allocated 2,308 FTE positions to the eligibility determination
process. Of these positions, 520 were vacant for a vacancy rate of approximately 23
percent.
As shown in Table 4 (see page 8), it cost approximately $97 million in fiscal year 2011
to operate the eligibility process, of which approximately $69.4 million was transferred
to DES. The monies used to support this program consist of federal and state monies.
1 See footnote 1, page 3.
2 Under federal regulations, AHCCCS has no more than 45 days to either approve or deny an applicant for most eligibility
programs and must redetermine eligibility at least every 12 months.
page 8
State of Arizona
1 The Division of Member Services’ (Division) expenditures include only the Division’s direct costs for eligibility-related functions. According to
AHCCCS, it is not required to allocate overhead costs to various administrative functions because it is the single state agency for Medicaid
and the Children’s Health Insurance Program. Therefore, it claims such costs in accordance with CMS regulations. The transfers to DES
include DES’ indirect costs because it performs eligibility determinations for several different federal programs and has a cost allocation plan
to allocate indirect costs to various programs.
Source: Auditor General staff analysis of AHCCCS-prepared eligibility determination expenditures and transfers to DES for fiscal year 2011.
Table 4: Division of Member Services Eligibility Determination
Expenditures by Office and Transfers to DES1
Fiscal Year 2011
(Unaudited)
Federal State Total
Division of Member Services offices:
Field Operations Administration $ 11,908,074 $ 7,375,819 $ 19,283,893
Quality Compliance Administration 749,358 553,068 1,302,426
Member Database Management Administration 832,359 282,511 1,114,870
Office of Automation 736,293 269,006 1,005,299
Administrative Services Office 843,284 3,183,250 4,026,534
Program Support Administration 471,740 353,737 825,477
Total Division of Member Services eligibility determination expenditures 15,541,108 12,017,391 27,558,499
Transfers to the Department of Economic Security 31,542,570 37,832,712 69,375,282
Total eligibility determination expenditures and transfers $ 47,083,678 $ 49,850,103 $ 96,933,781
AHCCCS should take some additional
actions to strengthen the eligibility
determination process
FINDING 1
page 9
5.92 percent of individual eligibility determinations
are at risk for processing errors
Although AHCCCS and DES appropriately determined the eligibility for
approximately 94 percent of individuals who applied for Medicaid services,
5.92 percent of eligibility determination cases are at risk for processing
errors.1,2 Auditors reviewed a representative sample of 279 individuals enrolled
in either an AHCCCS Acute
Care or Arizona Long Term Care
System (ALTCS) program in
April 2011 to determine the rate
at which caseworkers were
making processing errors (see
Appendix B, pages b-i through
b-ix, for additional information
on the sample). To assess the
accuracy of the eligibility
determination for the 279
enrolled members, auditors
focused their work on six key
eligibility criteria established in
AHCCCS’ state plan and
eligibility policies that are
consistent with federal regulation
(see textbox).
Auditors’ review determined that AHCCCS and DES correctly processed
approximately 94 percent of eligibility determinations. However, for 16 of the
279 eligibility determinations, auditors identified processing errors. Auditors
calculated a processing error rate for the population of 5.92 percent. These
processing errors involved problems related to correctly calculating and
verifying income or resources, both critical components of the eligibility
determination process. Auditors did not identify any problems with other
1 As indicated in the Introduction (see page 5), AHCCCS has an intergovernmental agreement with DES to
conduct Medicaid eligibility because DES performs this function for other programs, such as the Supplemental
Nutrition Assistance Program and Temporary Assistance for Needy Families program.
2 The 5.92 percent is a weighted error rate calculated by auditors. See Appendix B, pages b-iii through b-iv, for
more information on this calculation.
The Arizona Health Care
Cost Containment System
(AHCCCS) should take
some additional steps to
further strengthen the
eligibility determination
process. AHCCCS and the
Department of Economic
Security (DES) accurately
determined eligibility for
almost all Medicaid
applicants. However,
auditors calculated that
5.92 percent of the
eligibility determinations
are at risk for processing
errors, and 1.11 percent of
the eligibility determinations
are at risk for being
incorrect. Therefore,
auditors estimated that
AHCCCS is paying
between approximately
$3.5 and $4.8 million in
monthly capitation
payments for enrolled but
ineligible members. To help
further ensure that
Medicaid monies are spent
only on those members
who meet eligibility
requirements, AHCCCS
should develop and
implement a corrective
action plan to address the
errors identified. In addition,
AHCCCS should increase
its use of available
electronic means for
verifying citizenship to
further enhance its eligibility
determination process.
Office of the Auditor General
Six key eligibility criteria reviewed
• United States citizenship or qualified alien
status
• Valid social security number
• Arizona residency
• Income requirements
• Signed application
• Resource limits for the Arizona Long Term
Care System (ALTCS)
Source: Auditor General staff analysis of AHCCCS’ State
Plan and eligibility policies.
page 10
State of Arizona
eligibility criteria, such as citizenship or residency, for the 279 eligibility determination
cases reviewed. Processing errors specifically included the following:
• Lack of sufficient documentation showing caseworkers verified income—
For 9 of the 16 eligibility determinations where auditors identified processing
errors, AHCCCS and/or DES caseworkers did not verify or document,
as required, the individual’s or family’s income. AHCCCS and DES
policies require that caseworkers document verification of reported
income (see textbox). Based on auditors’ review of the 279 cases,
caseworkers appropriately used a variety of sources including pay
stubs, third-party databases, and employer statements to verify
and document reported income (see Table 5). However, auditors
found that policies for income verification and documentation were
not followed for 9 of the cases reviewed, resulting in insufficient
verification and documentation of income. For example:
Table 5: Income Verification Sources Used by AHCCCS and
DES to Confirm Reported Income
April 2011
1 Auditors’ review of 279 eligibility determinations resulted in 300 instances of income verification because
some households had multiple income sources. In addition, as noted in the bulleted paragraph above,
auditors also found that policies for income verification and documentation were not followed for 9 of
the cases reviewed.
2 Although caseworkers relied on a self-declaration of reported income for these cases, caseworkers may
also have checked other verification sources such as The Work Number or the State Verification
Exchange System in an effort to verify reported income.
Source: Auditor General staff analysis of 279 eligibility determinations for members active in an AHCCCS
Acute Care or ALTCS program in April 2011.
e
Income requirements
Income requirements vary by age and other
factors such as whether the applicant is
pregnant, and are based on federally
determined poverty guidelines. For example,
an individual can make up to $908 per month
and still qualify, while a family of 4 can make
up to $1,863 per month. For each additional
family member, the income limit increases by
$318 per month.
Source: Auditor General staff analysis of information
provided by AHCCCS.
Verification Type Example(s)
Form of
Electronic
Verification
(Yes or No)
Number of
Verifications
(Percentage
of Total)1
Electronic
verification
Third-party databases such
as The Work Number (see
pages 14-15)
Yes 111 (37.0%)
Third-party
documentation
Pay stubs, bank
statements, and child
support statements
showing income amounts
No 100 (33.3%)
Informal employer
verification
Note from employer
corroborating applicant’s
reported income
No 62 (20.7%)
Self-declaration2 Applicant’s self-reported
income information
No 27 (9.0%)
page 11
Office of the Auditor General
◦ In one eligibility determination, the caseworker relied on the applicant’s
statement of reported income and did not verify the reported income through
an employer’s statement or pay stubs. AHCCCS subsequently determined
that the applicant’s monthly income was approximately $380 more than the
amount the caseworker used to make the determination.
◦ In another eligibility determination, the caseworker did not appropriately
obtain income verification documentation. AHCCCS subsequently obtained
verification showing that the monthly income allocated to the applicant
should have been approximately $650 higher.
In these two examples, the processing errors were not of sufficient magnitude to
change the determination that these applicants were eligible for Medicaid.
Nonetheless, it is important to verify and document income as required, because
the amount of income is a key factor for determining individual and/or family
eligibility for Medicaid.
• Income or resource miscalculations—For 7 of the 16 eligibility determinations
where auditors identified processing errors, AHCCCS and/or DES caseworkers
miscalculated the individual’s or family’s amount of income or resources. All
applicants must meet income requirements and ALTCS applicants must also
meet resource requirements (see
textbox). Therefore, correctly determining
income and resources is an essential
component of the eligibility determination
process. Using various documents
obtained during the application process,
caseworkers must often calculate and
input income and resource information
into a computerized eligibility system.
For example, income eligibility is
determined based on monthly income,
and caseworkers must manually convert
weekly and biweekly income amounts
into monthly income amounts. The
caseworker must then enter the
calculated amount into the eligibility
system so that eligibility can be
determined.
The calculations that caseworkers must make can lead to mistakes. For example,
in one determination, the caseworker incorrectly input an applicant’s monthly
income into the computer system, entering approximately $1,065 per month less
than should have been entered. In another determination, the caseworker did not
include approximately $550 in social security survivor’s benefits in an applicant’s
monthly income calculation. Although only one of the seven income/resource
ALTCS resource requirements
For single applicants in the ALTCS program, countable resources
cannot exceed $2,000.
Countable resources include:
• Financial accounts, such as savings accounts
• Life insurance or burial funds in excess of $1,500
• Cash, stocks, and bonds
Not countable resources include:
• The home that the applicant lives in
• One vehicle
• Household and personal belongings
Source: Auditor General staff analysis of ALTCS eligibility policies.
page 12
State of Arizona
miscalculations auditors identified resulted in an incorrect Medicaid benefits
determination, it is important to correctly calculate and enter income and
resource information because these amounts are key determinants in whether
an individual is eligible for Medicaid.
Processing errors create potential for incorrect eligibility
determinations
For 13 of the 16 determinations with processing errors, the errors that auditors
identified were not of sufficient magnitude to cause incorrect eligibility determinations,
such as certifying applicants as eligible when they actually were not. However, for
three of the sampled determinations, the processing errors were of sufficient
magnitude to result in incorrect determinations. In these instances, caseworkers did
not properly verify or correctly calculate the income amount used to make the
eligibility determination. In the first case, there was no documentation in the member’s
case file to support the eligibility determination. AHCCCS later determined that the
member was not eligible because the income verification information that AHCCCS
later found showed that the member exceeded the income limit. In October 2011,
this individual was not reapproved for Medicaid services because she did not meet
the income requirements. In the second case, the caseworker incorrectly entered net
earnings instead of gross earnings, and therefore the income reported was $92 less
than it should have been. If the amount had been entered correctly, the member’s
income would have been approximately $62 over the income limit. In the third case,
AHCCCS was not able to provide income verification as required for the period under
review and absent this verification, the applicant should not have been approved for
Medicaid benefits. As a result, AHCCCS inappropriately paid approximately $2,359
in capitation payments for these three members over a 7-month period.1
Although the incorrect eligibility determinations resulting in erroneous payments were
small within the actual sample of cases reviewed, they become more important when
projected across all Arizona Medicaid eligibility determinations performed by
AHCCCS and DES. Specifically, auditors calculated an incorrect eligibility
determination rate of 1.11 percent for the acute care and long-term care programs
tested. Therefore, of the approximately $414 million in monthly capitation payments
that AHCCCS makes for its members in the programs tested, auditors estimated that
AHCCCS is paying its health plans between approximately $3.5 and $4.8 million in
monthly capitation payments for enrolled but ineligible members (see Appendix B,
pages b-v through b-vi, for auditors’ calculation of this range).2
1 The capitation dollars paid in error were calculated based on monthly capitation payments that AHCCCS made for
each member between the date of the erroneous determination and the end of the period for which auditors had
capitation payment data—July 31, 2011. These capitation dollars do not include payments AHCCCS made for
children’s rehabilitation or mental health services.
2 Auditors’ estimates of the capitation payment amounts made for enrolled but ineligible members do not include
payments made for children’s rehabilitation or mental health services.
Three of the 16
processing errors
resulted in an incorrect
eligibility determination.
page 13
Office of the Auditor General
Most error types consistent with federal review
The types of errors auditors found are consistent with those identified by the most
recent federally required Payment Error Rate Measurement (PERM) review (see
textbox). Arizona was 1 of 17 states that underwent the PERM review in federal fiscal
year 2008. This PERM review found that Arizona’s eligibility error rate was 2.2 percent,
which was lower than the national eligibility error rate of 6.7
percent reported by the review. Arizona had the third
lowest eligibility error rate of the 17 states under review.
The federal fiscal year 2008 PERM review error rate for
Arizona of 2.2 percent included eligibility errors in three
categories: 1) incorrect eligibility determinations due to
members being approved for Medicaid, but not meeting
the requirements, 2) members for whom eligibility could
not be determined because the case record contained
insufficient documentation, and 3) cases where the State
should have paid a greater percentage of the member’s
institutional care costs when these are shared costs.
Auditors similarly identified eligibility errors in the first two
categories for the sample of 279 determination cases
reviewed, but did not test these cases for the third
category. Arizona received another PERM review in
federal fiscal year 2011, and CMS is expected to notify
Arizona of the results by November 2012.
AHCCCS should implement plan to
reduce errors made during eligibility
determinations
AHCCCS should take several steps, through the development of a corrective action
plan, to correct the kinds of errors this audit identified, as well as to minimize their
frequency going forward. Minimizing errors is important because AHCCCS cannot
recover monies paid for someone it incorrectly determined eligible unless the
information used to make the determination is proven to be fraudulent. Specifically,
AHCCCS management explained that the State is not able to recover any of the
monies inappropriately paid to the health plans and/or providers for these members
because these monies were provided to the health plans and providers to cover the
healthcare costs of members who the State had determined eligible and enrolled in
one of its contracted health plans. However, according to AHCCCS policy, if a
caseworker suspects fraud or abuse during the application process, they are
responsible for making a referral to AHCCCS’ Office of the Inspector General. The
Office investigates providers and members who are suspected of fraud, recovers
Payment Error Rate Measurement (PERM)
review
The U.S. Department of Health and Human Services,
Centers for Medicare and Medicaid Services (CMS)
requires state Medicaid agencies to undergo a PERM
review every 3 years to evaluate the accuracy of
Medicaid eligibility determinations, and managed care
and fee-for-service payments. A contractor determines
the sample size on behalf of the federal government
based on the eligibility error rate found in the previous
PERM review, and then AHCCCS quality compliance
staff conduct the PERM review on behalf of the federal
government. For the eligibility determination portion of
the PERM review, AHCCCS staff review all
documentation that the caseworker used to approve
or deny Medicaid eligibility for each case in the
sample to confirm that the eligibility determination was
correct.
Source: Auditor General staff review of PERM guidance documents
and Arizona’s federal fiscal year 2008 PERM report.
page 14
State of Arizona
overpayments, issues administrative sanctions, and refers cases for criminal
prosecution.
AHCCCS previously implemented actions to correct eligibility determination errors
identified during the federal fiscal year 2008 PERM review. Its corrective actions
focused on implementing a new electronic file imaging system, rewriting areas of
unclear policy, and issuing policy training bulletins to caseworkers specific to the
errors found. Although both AHCCCS and DES took these steps, given the
complicated nature of determining and verifying income, such as ensuring that all
sources and amounts of income have been identified, appropriately calculated, and
verified, errors are still being made. Therefore, AHCCCS should develop a corrective
action plan that will help ensure that it and DES correct the types of income and
resource verification and calculation errors identified in this audit as well as minimize
their frequency going forward. Similar to corrective action strategies implemented by
other states to reduce eligibility errors, AHCCCS’ plan should include additional
caseworker training, particularly in areas determined to be error-prone, and
assessing whether it needs to clarify its income and resource policies. Additionally,
the plan should focus on enhancing the supervisory review of income and resource
requirements for Medicaid determinations. One way this could be done would be to
assess whether new or experienced caseworkers are making errors and further
target supervisory review.
AHCCCS should increase its use of electronic matching
to verify citizenship requirements
To further enhance its eligibility determination process, AHCCCS should increase its
use of available electronic means for verifying citizenship. Caseworkers consistently
verified social security numbers using electronic matching, but not income or
citizenship. AHCCCS and DES provide caseworkers with electronic means for
verifying social security numbers, income, and citizenship (see textbox, page 15).
Caseworkers used the electronic interface with the Social Security Administration to
verify the social security number of the applicant in all 279 eligibility determinations
auditors reviewed.
AHCCCS’ and DES’ use of electronic matches to verify reported income was either
less successful or less frequent. Specifically, as shown in Table 5 (see page 10),
caseworkers were able to use electronic matches, such as The Work Number, to
verify income 37 percent of the time for the 279 determinations auditors reviewed.
Although auditors found that DES caseworkers used electronic means in additional
cases to verify reported income, these electronic matches did not produce the
information needed to verify the reported income. Additionally, for some of the
eligibility determinations performed by DES, it did not use electronic means to verify
the reported income. As a result, the remaining income verifications were completed
through other means, such as pay stubs, bank statements, employer statements,
AHCCCS should develop a
corrective action plan that
includes additional training,
policy clarification, and
enhanced supervisory
review.
page 15
Office of the Auditor General
and applicant’s self-reporting of income. However,
according to a DES policy, as of June 2011, information
from the Work Number is viewed through DES’ eligibility
system for new and renewal applicants age 16 and
older who have a valid social security number.
According to a DES official, this change will increase
DES’ use of this electronically verified income
information during the eligibility process.1
Finally, the use of electronic verification for citizenship
was also less frequent. As shown in Table 6 (see page
16), caseworkers verified citizenship with hard-copy
documents, such as birth certificates, instead of
electronic matches such as the AHCCCS Citizenship
Verification System, for 72 percent of the 279
determinations. To further enhance its eligibility
determination process, AHCCCS should ensure that it
and DES make greater use of the available electronic
means for verifying citizenship. Although auditors did
not identify problems with citizenship for the 279
eligibility determinations reviewed, electronic verification
would offer the opportunity to streamline the eligibility
process. Specifically, verifying citizenship electronically
may require interagency cooperation to share data and
costs to improve information technology, but it reduces
unnecessary paperwork for families without sacrificing
accuracy and eases the administrative burden on the
agency.2,3,4
1 Auditors’ review of a representative sample of the 279 individuals enrolled in either an AHCCCS Acute Care or ALTCS
program did not identify that AHCCCS needed to increase its use of electronic matches for income verification.
AHCCCS conducts Medicaid eligibility determinations for the ALTCS program. As indicated in the Introduction (see
page 2), this program is for individuals who are elderly or disabled. Auditors’ review of the ALTCS determinations found
these individuals had unearned income, such as social security income, and not earned income. In addition, AHCCCS
used an electronic match to verify unearned income for most determinations.
2 Kaiser Commission on Medicaid and the Uninsured. (2011). Holding steady, looking ahead: Annual findings of a
50-state survey of eligibility rules, enrollment, and renewal procedures, and cost sharing practices in Medicaid and CHIP,
2010-2011. Washington, DC: The Henry J. Kaiser Family Foundation.
3 Cohen, A., Dutton, M., Griffin, K., Woods, G., & Manatt Health Solutions. (2008). Streamlining renewal in Medicaid and
SCHIP: Strategies from other states and lessons for New York. New York, NY: Author.
4 Dorn, S. (2009). Express lane eligibility and beyond: How automated enrollment can help eligible children receive
Medicaid and Chip: A catalog of state policy options. Washington, DC: Urban Institute, Health Policy Center.
Electronic verification sources
Income verification:
• State Verification Exchange System (SVES)—This
system obtains information from other sources such
as Unemployment Insurance and the Social Security
Administration.
• The Work Number—An automated online
employment and income verification system which
allows DES or AHCCCS to retrieve employment and/
or payroll verification from participating employers.
Social security number verification:
• Wire Third Party Query System (WTPY)—Interface
with the Social Security Administration to verify the
applicant’s social security number.
Citizenship and qualified alien status verification:
• AHCCCS Citizenship Verification System—A Web-based
birth records match with Arizona Vital Records
to verify birth records for individuals born in Arizona
1950 and later.
• Systematic Alien Verification for Entitlements
(SAVE) —Verifies the authenticity of the document the
applicant provided to prove his/her alien status.
Source: Auditor General staff summary of AHCCCS and DES eligibility
policies and procedures.
page 16
State of Arizona
Electronic verification will be required by the Affordable
Care Act and is being increasingly adopted by other
states to verify citizenship
The 2010 Federal Patient Protection and Affordable Care Act (Act) directs states to
expand their use of electronic verification for Medicaid eligibility by 2014.1 For
example, under the Act’s proposed rules, Medicaid eligibility verification will include
interfaces with agencies such as the Social Security Administration for citizenship
verification, and immigration status through the Department of Homeland Security.
Although not all of the federal rules are promulgated nor are all of the processes in
place within the federal government for Arizona to fully implement the electronic
Medicaid eligibility verification requirements of the Act, Arizona has begun planning
for implementation. Specifically, in February 2012, according to an AHCCCS official,
the State began developing some of the requirements that will allow it to operate a
Health Benefit Exchange.2 This Exchange is intended to enable, among other things,
the electronic verification of Medicaid eligibility requirements and make determinations
as close to real time as possible.
Many states have already begun to use the electronic verification for citizenship that
will be required by the Act. According to Kaiser Foundation statistics, 40 states have
established an electronic data match with the federal Social Security Administration
to verify citizenship status for parents applying for Medicaid as of January 2012.3 This
1 The constitutionality of portions of the Patient Protection and Affordable Care Act is under review by the United States
Supreme Court. The case involves numerous parties. Oral arguments were heard in late March 2012 with a ruling
expected sometime in late June 2012.
2 According to the Act, no later than January 1, 2014, each state shall establish an American Health Benefit Exchange,
administered by a governmental or nonprofit entity established in the state, that makes qualified health plans available
to qualified individuals and employers.
3 Kaiser Foundation updates the states’ use of the Social Security Administration Data Match as part of its dynamic Web
Site. See http://www.statehealthfacts.org/comparemaptable.jsp?ind=899&cat=4
Table 6: Citizen and Qualified Alien Status Verification Sources
Used by AHCCCS and DES
April 2011
Source: Auditor General staff analysis of 279 eligibility determinations for members active in an
AHCCCS Acute Care or ALTCS program in April 2011.
Verification Type Example(s)
Percentage
of Total
(279)
Third-party
documentation
Birth certificate, passport,
naturalization certificate
72%
Electronic
verification
AHCCCS Citizenship
Verification System,
Systematic Alien Verification
for Entitlements
28
page 17
Office of the Auditor General
represents an increase of 12 states since January 2011.1 Although Arizona has not
established this match, most states, including California and Texas, have established
this match. State experience with this option found that this match accurately verified
citizenship in 94 percent of cases, while easing the state’s administrative workload
and eliminating unnecessary paperwork for applicants. AHCCCS indicated that it
has not established an interface with the Social Security Administration to verify
citizenship because of concerns about having to grant an applicant temporary
eligibility in cases where the social security number could not be verified through the
match.2
Recommendations:
1.1 AHCCCS should develop a corrective action plan that will help ensure that it
and DES correct the types of income and resource verification and calculation
errors identified in this audit as well as to minimize their frequency going
forward. This plan should include:
a. Providing additional caseworker training, particularly in areas determined
to be error-prone;
b. Assessing whether it needs to clarify its income and resource policies;
and
c. Enhancing the supervisory review of income and resource requirements
for Medicaid determinations.
1.2 AHCCCS should ensure that it and DES make greater use of the available
electronic means for verifying citizenship, such as the AHCCCS Citizenship
Verification System.
1 Kaiser Commission on Medicaid and the Uninsured. (2011). Holding steady, looking ahead: Annual findings of a
50-state survey of eligibility rules, enrollment, and renewal procedures, and cost sharing practices in Medicaid and CHIP,
2010-2011. Washington, DC: The Henry J. Kaiser Family Foundation.
2 Section 1902 of the Social Security Act requires that if the citizenship information submitted by the state Medicaid
agency cannot be verified by the Social Security Administration, the applicant will be granted 90 days of temporary
Medicaid eligibility, during which time the applicant would have to provide documentation to verify citizenship status.
After 90 days, if the applicant’s citizenship cannot be resolved, the state Medicaid agency would remove the individual
from the rolls within 30 days.
Most states have
established the electronic
verification for citizenship
that is required by the
Act.
page 18
State of Arizona
APPENDIX A Methodology
page a-i
Auditors used the following specific methods to meet the audit’s objectives:
• To gain an understanding about the state and federal requirements for
performing eligibility determination, auditors reviewed the Code of Federal
Regulations, the Office of Management and Budget Circular A-133
Compliance Supplement Catalog of Federal Domestic Assistance 93.778
Medical Assistance Program, AHCCCS’ State Plan for Medicaid, the
intergovernmental agreement between AHCCCS and DES, and AHCCCS’
and DES’ eligibility policies and procedures; and conducted interviews
with AHCCCS and DES staff.1
• To determine the rate at which AHCCCS and DES caseworkers were
making processing errors on Medicaid eligibility determinations, auditors
met with agency staff to identify relevant eligibility and payment data
systems, reviewed systems information, and ultimately obtained eligibility
records and payment data for all members active in April 2011. Auditors
used the data to select a representative sample of 279 eligibility
determinations. Auditors used this sample to assess whether AHCCCS
and DES were appropriately determining eligibility, calculate an error rate,
and construct an estimate of the monthly capitation payments made for
enrolled but ineligible members (see Appendix B, pages b-i through b-ix,
for technical information on the sample, error rates, and projection
methodology).
• Auditors’ work on internal controls involved reviewing AHCCCS’ and DES’
written policies and procedures for guiding eligibility work, including its
quality control processes. For example, auditors observed a case review
conducted by AHCCCS’ Quality Compliance Administration; reviewed the
federal fiscal year 2008 Arizona Payment Error Rate Measurement (PERM)
findings, corrective action plan, and associated manuals; interviewed
consultants responsible for providing states with sampling requirements;
and reviewed the Medicaid Eligibility Quality Control review process. In
addition, auditors reviewed information technology change management
controls for both AHCCCS’ and DES’ eligibility systems and the
reconciliation process between these eligibility systems and AHCCCS’
Prepaid Medical Management Information System.
1 As indicated in the Introduction section (see pages 1 through 8), AHCCCS has an intergovernmental
agreement with DES to conduct Medicaid eligibility because DES performs this function for other programs,
such as the Supplemental Nutrition Assistance Program and Temporary Assistance for Needy Families
program.
This appendix provides
information on the methods
auditors used to meet the
audit objectives.
This performance audit was
conducted 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.
The Auditor General and
staff express appreciation
to the Arizona Health Care
Cost Containment System
(AHCCCS) Director and his
staff and the Department of
Economic Security (DES)
Director and his staff for
their cooperation and
assistance throughout the
audit.
Office of the Auditor General
page a-ii
State of Arizona
• To ensure the eligibility and payment data was generally complete and accurate
for the purposes of selecting a representative sample and developing estimates
of capitation payments made for enrolled but ineligible members, auditors
conducted various tests such as comparing number of records and capitation
data received from AHCCCS to published population figures and capitation
rates. Auditors determined the data was generally complete and accurate for
the audit’s purposes.
• To evaluate AHCCCS’ use of electronic verification during the eligibility
determination process and efforts to prepare for federal healthcare reform,
auditors reviewed the 2010 Patient Protection and Affordable Care Act (Act);
U.S. Department of Health Services, Centers for Medicare and Medicaid
Services’ proposed rules associated with implementation of the Act; electronic
data matches with the U.S. Social Security Administration; and information
prepared by AHCCCS staff involved with the Act’s implementation. In addition,
auditors obtained information on other states’ use of electronic verification from
the Henry J. Kaiser Family Foundation and Urban Institute.
• The information used to develop the report’s Introduction included AHCCCS
documents published on its Web site, such as information about court decisions;
AHCCCS’ fiscal year 2010 and 2011 audited financial statements; and
information from AHCCCS and DES administrators and staff, including internal
staffing reports.
Sampling, error rate estimate, and
projection methodology
APPENDIX B
Technical Methods
page b-i
Process for selecting the sample
To select a representative sample, auditors completed the following activities:
• Identifying the population—Because auditors focused on assessing
eligibility determinations, auditors identified the original population from
which to draw a sample as any member with an eligibility record in a
single month. Auditors then obtained the eligibility records for each mem-ber
enrolled in an AHCCCS healthcare program at any point during April
2011. For each member enrolled, auditors obtained all of the eligibility
records associated with that member during June 1, 2010 through July
31, 2011. Some members had more than one eligibility record because
of changes to their eligibility that may have qualified them for a different
AHCCCS program. Each change results in a new eligibility record. An
eligibility record includes information such as the eligibility code for the
specific program the member is enrolled in, and the beginning and end-ing
dates of the enrollment. AHCCCS members receive medical services
under three primary programs: KidsCare and Arizona’s two Medicaid
programs—Arizona Long Term Care System (ALTCS) and Acute Care
(see Introduction, page 2, for more information on these three programs).
The Acute Care program has several subprograms that cover specific
populations, such as low-income families with children, or pregnant
women. Auditors grouped members’ eligibility records by programs/sub-programs
using the eli-gibility
codes includ-ed
in the data. This
resulted in a popula-tion
of 2,120,900 eligi-bility
records for
1,495,203 members
(see Population and
sampling frame text-box).
• Identifying the sampling frame—Auditors narrowed the 2,120,900
records in the population to only those members with active records in the
ALTCS program and the three largest Acute Care subprograms: Families
with Children, S.O.B.R.A. Child, and AHCCCS Care. This population was
identified as the audit sampling frame and although it was smaller than
the original population, it still represented the majority of AHCCCS records
This appendix provides
information on the methods
auditors used to select
their sample, estimate
eligibility processing and
eligibility determination
error rates, and develop an
estimate of capitation
payments the Arizona
Health Care Cost
Containment System
(AHCCCS) is making to its
contracted health plans for
ineligible members.
Contents:
• Sample selection (pages
b-i through b-iii);
• Eligibility processing error
rate estimates (pages b-iii
through b-iv);
• Eligibility determination
error rate estimate (page
b-iv);
• Incorrect capitation pay-ment
estimate (pages
b-v through b-vi);
• Sample demographics
(page b-vii); and
• Glossary of statistical
terms used in this appen-dix
(pages b-viii through
b-ix).
Office of the Auditor General
Population and sampling frame
Population
Sampling
Frame
Percentage
of Population
Records 2,120,900 1,460,094 69 %
Members 1,495,203 1,183,559 79 %
page b-ii
State of Arizona
and members (see textbox, page b-i). The sampling frame excluded records for
members who fell outside the scope of the audit, such as members whose
eligibility is determined by the Foster Care Program, automatic re-determinations
(6-month extensions), and determinations made in part by outside entities, such
as the U.S. Social Security Administration for individuals who are aged, blind, or
disabled and qualify for Supplemental Security Income.
• Selecting a stratified random sample—Auditors designed an embedded
stratified random sample for the audit. The procedures and processes for
determining eligibility for the AHCCCS programs auditors tested are the same;
however, the level of income an individual is allowed to have varies, and the
ALTCS program has a resource limit (see Finding 1, textbox, page 11). Auditors
randomly selected 280 members stratified first by program and then by
subprogram.1 The first stratification differentiated Acute Care and ALTCS. Acute
Care represented the majority of records in the population while ALTCS
represented less than 5 percent of all records in the population, but a significant
amount of the capitation payments. Auditors then further stratified Acute Care
into three subprograms that accounted for more than 66 percent of all Acute
Care records. The member’s active record was used as a surrogate for the
member and could be selected only one time. For the Acute Care program,
auditors selected a random sample proportionate to the size of 250 members’
eligibility records from the sampling frame, stratified by program, which resulted
in sampling 116 members from Families with Children, 82 members from
S.O.B.R.A. Child, and 52 members from AHCCCS
Care. For the ALTCS program no further stratification
was necessary and auditors randomly sampled 30
members.2 This stratum is oversampled when
combined with the Acute Care strata, but the sample
proportions were adjusted to account for the
oversampling post hoc. Auditors wanted to ensure that
the sample size for all strata were large enough to
draw conclusions and make projections, if necessary,
by strata to the members in the population (see
stratified random sample textbox).
• Adjusting the population for projection purposes—Auditors adjusted the
sampling frame to account for the oversampling and to exclude any members
not subject to testing, e.g. inactive members and foster care children. These
post hoc adjustments resulted in a final population of 1,169,351 active members,
1,394,500 eligibility records, and minor changes to the strata proportions used
in weighting (see Adjusted population textbox, page b-iii). The proportions used
in weighting error rates and projections are representative of the final adjusted
1 Auditors’ initial sample size was 280; however, one member record was dropped from the initial sample of 30 for ATLCS
because it was inactive. Auditors did not replace that item as the stratum was oversampled and the loss of one case
was negligible to the sample.
2 See footnote 1.
Stratified random sample
Sample
Program Records Percentage Size
Families with Children 653,073 46.4 % 116
S.O.B.R.A. Child 459,486 32.7 82
AHCCCS Care 293,524 20.9 52
Total Acute Care 1,406,083 100.00 %
ALTCS 54,011 29
Total 1,460,094 279
page b-iii
Office of the Auditor General
population. The sample has a 95 percent
level of confidence with a sample precision
of plus/minus 3 percent.
• Ensuring representativeness of the
sample—Auditors compared key
demographics of the sampled members’
records, such as age and geographical
location, to the sampling frame population
to ensure that the sample was
representative of the sampling frame
population. This was done with the initial
sampling frame and then again for the adjusted population. There were no
statistical differences between the sample and the initial sampling frame or the
sample and the final population (see Sample demographics, page b-vii).
Steps for estimating processing errors
Auditors calculated a processing error rate. Processing errors are mistakes made
during the eligibility determination process, such as a caseworker not documenting
income verification or miscalculating the applicant’s income (see Finding 1, pages 9
through 17). To determine the rate at which processing errors were likely to occur in
the population, auditors estimated a weighted processing error rate based on sample
results. A weighted error rate is commonly used to minimize any design effects and
reduce sampling error. The weighted processing error rate also accounts for any
variation among the strata that could unwittingly bias the estimated rate of processing
error. Auditors took the following steps to estimate the rate at which processing errors
were likely to occur in the population: (Also see Processing error rates textbox, page
b-iv.)
• Determining the unweighted processing error rate—This rate represents the
number of processing errors for a stratum as a proportion of the sample size. For
example, the unweighted processing error rate for the AHCCCS Care program is
4/52 = 7.69 percent.
• Calculating the weighted processing error rate—This rate is calculated by
multiplying the unweighted processing error rate for a stratum by the proportion
of the population represented by that stratum. For example, the weighted
processing error rate for the AHCCCS Care program is: 0.0769 x 0.2055 = 1.58
percent.
• Estimating a total weighted processing error rate—This rate is calculated by
summing the weighted error rates of each stratum. This calculation resulted in a
1 The adjusted population is the sampling frame minus inactive records and other
records not subject to testing (see page b-ii).
Adjusted population1
Population
Sampling
Frame
Adjusted
Population
Adjusted
Population as
a Percentage
of Population
Records 2,120,900 1,460,094 1,394,500 66%
Members 1,495,203 1,183,559 1,169,351 78%
page b-iv
State of Arizona
weighted stratified error rate of 5.92 percent that can be projected to the
members in the population.1
Process for calculating an incorrect eligibility
determination rate
For those members with processing errors resulting in an incorrect eligibility
determination, auditors computed an incorrect eligibility determination rate (see
Finding 1, pages 9 through 17). The incorrect eligibility determination rate represents
the probability in the population of a member having at least one processing error on
their last determination that also resulted in an incorrect eligibility determination. This
rate was calculated as follows:
• Determining the percentage of processing errors resulting in incorrect
eligibility determinations—From the sample, auditors found that 3 of the 16
members’ processing errors resulted in an incorrect eligibility determination
(see Finding 1, pages 9 through 17). Therefore, the percentage of processing
errors resulting in incorrect eligibility determinations is 3/16 = 0.1875, or 18.75
percent.
• Calculating the incorrect eligibility determination rate—This rate is determined
by multiplying the weighted processing error rate by the percentage of incorrect
eligibility determinations, or 0.0592 x 0.1875 = 0.0111, or 1.11 percent.
1 T-tests of difference of proportions (alpha=.05) showed that there were no statistical differences between the error rates
by strata.
Processing error rates
Percentage of
Adjusted
Records1
Adjusted
Records1
Sample
Size
Processing
Errors
Processing Error Rate
Program Unweighted Weighted
Families with Children 628,591 45.08% 116 7 6.03% 2.72 %
S.O.B.R.A. Child 425,584 30.52 82 4 4.88 1.49
AHCCCS Care 286,622 20.55 52 4 7.69 1.58
ALTCS 53,703 3.85 29 1 3.45 0.13
Total 1,394,500 100.00% 279 16 5.92 %
1 Population proportions were adjusted to exclude inactive and other records not subject to testing (see page b-ii).
page b-v
Office of the Auditor General
Steps for estimating incorrect capitation payments
Incorrect determinations can be tied directly to capitation payments made in error. To
construct an estimate of the monthly capitation payments made for enrolled but
ineligible members, auditors completed the following procedures: (Also see textbox,
page b-vi.)
• Projecting the number of incorrect eligibility determinations in the population
by stratum—Auditors estimated the number of members in the population that
would have incorrect eligibility determinations. This number is determined by:
◦ Multiplying the unduplicated number of members in the population used for
projection purposes (1,169,351) by the incorrect eligibility determination rate,
or 1,169,351 x 0.0111 = 12,980. This number represents the number of
members who are projected to be ineligible for Medicaid; and,
◦ Apportioning the number of ineligible members (12,980) among the strata by
multiplying 12,980 by the percentage of members in each stratum. For
example, for the AHCCCS Care program the projected number of members
with eligibility determination errors would be 12,980 x 0.2055 = 2,668.
• Determining an average capitation payment for each stratum—Auditors used
the capitation payments made for each member in the sample during the period
June 1, 2010 through July 31, 2011, to develop an average per member, per
month capitation payment for each stratum.1 Auditors did not include fee-for-service
dollars in this calculation because there was an insufficient number of
members in our sample with fee-for-service payments to develop a valid average
fee-for service payment by stratum. As reported in the Introduction (see page 2),
approximately 90 percent of AHCCCS members are enrolled in managed care
programs and their healthcare costs are covered through monthly capitation
payments made to contracted health plans. Capitation payment amounts also do
not include payments made for children’s rehabilitative or mental health services.
• Estimating the monthly incorrect capitation payment in the population—To
determine the estimated incorrect monthly capitation payment for the population,
auditors multiplied the average per member, per month capitation payment for
each stratum by the number of members in the stratum projected to have
incorrect determinations. This calculation for S.O.B.R.A. Child, for example, is
3,961 x $127.65 = $505,622. The calculation was made for each stratum and
summed to compute a total monthly estimate.
1 In addition to the monthly capitation payments, auditors’ analysis included some other types of capitation payments such
as lump sum payments for births that included the costs of prenatal care and the actual costs of the birth, and subsidies
for rural hospitals.
page b-vi
State of Arizona
• Developing a monthly incorrect capitation payment range—Using the
average monthly incorrect capitation payment for each stratum, auditors
estimated monthly incorrect capitation payments for the population. Auditors
also established upper and lower bounds around that estimate, meaning that
95 percent of the time the monthly amount of incorrect capitation payments
would fall within the established range.
1 Population proportions were adjusted to exclude inactive and other records not subject to testing (see page b-ii).
2 For the time period June 1, 2010 through July 31, 2011.
Projected estimate, lower, and upper monthly incorrect capitation payments
Average
Projected Per Member
Members Per Month
Percentage of
Members1
Incorrectly Capitation Monthly Incorrect Capitation Payment
Program Determined Payment2 Estimate Lower Bound Upper Bound
Families with Children 45.08 % 5,851 $ 175.40 $1,026,265 $ 887,421 $1,165,168
S.O.B.R.A. Child 30.52 3,961 127.65 505,622 449,851 561,353
AHCCCS Care 20.55 2,668 454.07 1,211,459 1,002,688 1,420,230
ALTCS 3.85 500 2,812.53 1,406,265 1,184,615 1,627,915
Total 100.00 % 12,980 $4,149,611 $3,524,575 $4,774,666
page b-vii
Office of the Auditor General
Sample demographics
Adjusted
Population1
Gender Sample
Male 46.66% 43.01%
Female 53.34 56.99
Total 100.00% 100.00%
Age
<1 0.11% 0.00%
1-13 41.97 37.28
14-44 44.65 46.59
45+ 13.27 16.13
Total 100.00% 100.00%
Race
Asian/Pacific Islander 1.81% 1.79%
Black 7.20 8.96
Cuban/Haitian 0.00 0.00
Caucasian/White 35.14 31.54
Hispanic 39.94 39.78
Native American 11.45 11.83
Other 0.20 0.00
Unknown/unspecified 4.26 6.09
Total 100.00% 100.00%
County
Apache 2.34% 1.79%
Cochise 1.90 2.51
Coconino 2.20 2.15
Gila 1.09 1.08
Graham 0.66 0.72
Greenlee 0.10 0.00
La Paz 0.33 0.00
Maricopa 54.67 60.93
Mohave 3.71 1.79
Navajo 2.94 1.43
Pima 15.05 13.98
Pinal 3.61 3.94
Santa Cruz 1.12 0.72
Yavapai 2.62 2.15
Yuma 3.88 2.87
Multiple 3.78 3.94
Total 100.00% 100.00%
1 The adjusted population is the sampling frame minus
inactive records or other records not subject to testing
(see page b-ii).
page b-viii
State of Arizona
Statistical glossary
Bias—Any source of systematic error in random probability sampling that is not
identified and accounted for when estimates to the population are made. Common
sources are disproportionate sampling, non-coverage of certain subpopulations, or
dependence among observed sampling units.
Confidence interval—An estimated range of values calculated from sample data to
include the unknown population value. The interval is constructed around a point
estimate from the sample such as a sample mean, and constructed for the
population at a specified level of confidence (95 percent or 99 percent) to assure that
95 or 99 percent of the time, the point estimate (mean) will fall within that interval in
the population. A confidence interval allows for random error when projecting a mean
value onto the population. A confidence level at alpha = 0.05 means that there is less
than a 5 percent chance that the interval will not contain the mean.
Confidence level—The specific probability of obtaining some result from the sample
that does not exist in the population. The confidence level establishes how confident
you are that you are not making a mistake in rejecting the null hypothesis. A 95
percent confidence level implies that 95 times out of 100 you would not be making
a mistake in rejecting the null hypothesis of no difference in the population.
Design effects—Bias created in statistical estimates when complex sampling
techniques are used. Methodological procedures have been developed to minimize
design effects, such as weighting and adjusting for measurable design effects.
Embedded stratified random sample—A complex sampling design where layers
of stratification are used to maximize sampling efficiency and minimize sampling
error.
Expected error rate—The amount of time you would expect to reject the null
hypothesis when it should not have been rejected. This number is also known as the
significance level for the test and is calculated as 100 minus the confidence level. By
definition, a 95 percent confidence level implies an expected error rate of 5 percent.
The expected error rate is also called the Type I or alpha level of the test of differences.
Most tests are conducted with an expected error rate of 5 percent or less, which
means that less than 5 percent of the time would you expect to incorrectly reject the
null.
Population—The group of items that the sample represents. The number of units in
the population is often denoted N.
Probability or random sample—A sample drawn from a population using a random
mechanism so that every element of the population has a known chance of ending
up in the sample. This type of sample can be generalized to the population so that
events observed in the sample can be projected to the population with confidence.
page b-ix
Office of the Auditor General
Sample size—The number of units in the sample, often denoted n.
Sampling error—Error created by unequal probability of selection in random
sampling. Unequal probabilities can be created by disproportionate sampling and
many other sources of sampling bias.
Sampling (or sample) frame—The target population from which the sample is
selected. Operationally, the sampling frame may differ from the overall population.
Significance level—The specific probability that the hypothesis test erroneously
rejects the null hypothesis when the null hypothesis is true. In most tests, the null
hypothesis is the hypothesis of no difference, i.e. that there is no difference between
the means or proportions of the two groups tested. When this hypothesis is rejected,
it means the two groups have the same mean or proportion, i.e. there is no difference
between the two groups. Hypotheses are tested at a specified level of significance
called the significance level. The significance level is usually set at 5 percent or less. It
is also known as the Type I or alpha level.
Stratified random sample—In a stratified random sample, subsets of the sampling
frame are selected separately, but randomly, from different categories or strata,
typically based upon a categorical variable of some kind such as program or
geography.
Stratum—In stratified random sampling, the sample is drawn separately from different
subsets of the population. Each such subset is called a stratum. The plural of stratum
is strata.
T-test—A small sample statistical test to determine if there is a significant difference
between two groups in the population based on the data from the sample. The t-test
should be used when samples are relatively small, < 30, and the distribution is known
to be approximately normal. The t-test can also be applied for samples of 30 to 120 as
it is a more precise test than its large-sample counterpart, the z-test. For sample sizes
>120, the t-distribution converges to the normal distribution, thus making the t-test
equivalent to the large sample test, the z-test. The test is usually set at a confidence
level (see below).
Upper and lower bounds of a confidence interval—Also known as confidence
limits, the upper and lower bounds are the values that define the range of the
confidence interval. The upper and lower bounds of a 95 percent confidence interval
are the 95 percent confidence limits and can be interpreted as the probability that the
estimate will fall in that interval. Ninety-five percent of the time, the interval will contain
the estimate and the estimate would not fall in the interval only 5 percent of the time.
AGENCY RESPONSE
Janice K. Brewer, Governor
Thomas J. Betlach, Director
801 East Jefferson, Phoenix, AZ 85034
PO Box 25520, Phoenix, AZ 85002
Phone: 602-417-4000
www.azahcccs.gov
Our first care is your health care
ARIZONA HEALTH CARE COST CONTAINMENT SYSTEM
May 29, 2012
Debra K. Davenport, CPA
Auditor General
Office of the Auditor General
2910 North 44th Street, Suite 410
Phoenix, AZ 85018
RE: Medicaid Eligibility Determination Performance Audit,
Revised Draft Report dated May 23, 2012
Dear Ms. Davenport:
Thank you for the opportunity to review and comment on the Arizona Health Care Cost Containment
System (AHCCCS) Medicaid Eligibility Determination Performance Audit. We appreciate the
professionalism and efforts of the audit team and believe that implementation of the findings will enhance
the efficiency and effectiveness of the AHCCCS eligibility process.
The past several years have proven challenging for State government and AHCCCS. Since the beginning
of the Great Recession, the AHCCCS Division of Member Services has faced a 39% reduction in staffing.
The Department of Economic Security (DES) has experienced a 15% reduction in Family Assistance
Administration staff. Despite these reductions, the counter-cyclical nature of the AHCCCS program has
resulted in an increased workload. In December 2007, approximately 198,000 determinations for
Medicaid benefits were made. As documented in your report, in December 2011 this number rose to
261,854, representing a 32% increase.
Your findings of a 1.11% error rate are commendable under these circumstances, particularly when
compared to the 2008 Payment Error Rate Measurement (PERM) national eligibility error rate of 6.7%.
The report recommends sound process improvements that should and will be made. However, the fact
that both agencies have carried on through these incredibly challenging times with an eligibility error rate
of only 1.11%, is a testament to the dedication and professionalism of the staff at AHCCCS and DES.
A recent employee survey of the AHCCCS Division of Member Services employees demonstrates the
level of dedication that exists:
􀂃􈌹 90% are proud to be an AHCCCS employee
􀂃􈌹 90% feel a sense of loyalty and commitment to AHCCCS
􀂃􈌹 94% understand clearly what is expected of them at work
􀂃􈌹 90% receive the assistance and guidance they need to do their job well
Given all that the agency has been through, it is important to highlight these inspiring results before
delving into the details of the report.
Ms. Debra K. Davenport, CPA
May 29, 2012
Page 2
The following responses address the recommendations proposed in the Revised Draft report:
Recommendation #1.1:
AHCCCS should develop a corrective action plan that will help ensure that it and DES correct the types
of income and resource verification and calculation errors identified in this audit, and to minimize their
frequency going forward. This plan should include:
a. Additional case worker training, particularly in areas determined to be error-prone;
b. An assessment of the need to clarify income and resource policies and;
c. Enhanced supervisory review of income and resource requirements for Medicaid determinations.
Response: The Finding of the Auditor General is agreed to, and the audit recommendation will be
implemented. In order to use limited resources most efficiently, (The AHCCCS Division of Member
Services has 39% fewer staff now as compared to 2007), these findings and recommendations will be
addressed in concert with the 2010 PERM Corrective Action Plan, to be developed and implemented late
this year upon receipt of final PERM results.
Recommendation #1.2:
AHCCCS should ensure that it and DES make greater use of the available electronic means for verifying
citizenship, such as the AHCCCS Citizenship Verification System.
Response: The Finding of the Auditor General is agreed to, and the audit recommendation will be
implemented. As noted in your report, we have been very proactive in implementing electronic
verification sources such as The Work Number and the AHCCCS Citizenship Verification System. Also,
as stated in the report, the use of additional electronic means, such as SSA citizenship data, is planned for
system development in support of the ACA. The AHCCCS Citizenship Verification System is limited to
information about individuals who were born in Arizona since 1950. Further, additional information, not
currently available in all cases (e.g. mother’s maiden name, place of birth), may be necessary to make
effective use of this system, particularly when working with applicants who have common names.
Although the transient nature of the population seeking benefits is another barrier to effective use of these
data, we will explore additional opportunities for eligibility staff to check this system prior to requesting
birth certificates or other forms of documentation. We propose to analyze options related to this
recommendation as part of the corrective action plan development described in the response to
recommendation #1.1.
Again, I would like to thank the Auditor General and staff for their time and effort in evaluating the
AHCCCS Medicaid Eligibility Determination program. We appreciate the professional approach of the
audit team as well as their cooperative attitude with AHCCCS staff.
Sincerely,
Thomas J. Betlach
Director
Performance Audit Division reports issued within the last 24 months
Future Performance Audit Division reports
Arizona State Parks Board
11-07 Department of Corrections—
Oversight of Security Operations
11-08 Department of Corrections—
Sunset Factors
11-09 Arizona Department of Veterans’
Services—Veterans’ Donations
and Military Family Relief Funds
11-10 Arizona Department of Veterans’
Services and Arizona Veterans’
Service Advisory Commission—
Sunset Factors
11-11 Arizona Board of Regents—
Tuition Setting for Arizona
Universities
11-12 Arizona Board of Regents—
Sunset Factors
11-13 Department of Fire, Building and
Life Safety
11-14 Arizona Game and Fish
Commission Heritage Fund
12-01 Arizona Health Care Cost
Containment System—
Coordination of Benefits
10-05 Arizona Department of Housing
10-06 Board of Chiropractic Examiners
10-07 Arizona Department of
Agriculture—Sunset Factors
10-08 Department of Corrections—
Prison Population Growth
10-L1 Office of Pest Management—
Regulation
10-09 Arizona Sports and Tourism
Authority
11-01 Department of Public Safety—
Followup on Specific
Recommendations from
Previous Audits and Sunset
Factors
11-02 Arizona State Board of Nursing
11-03 Arizona Department of Veterans’
Services—Fiduciary Program
11-04 Arizona Medical Board
11-05 Pinal County Transportation
Excise Tax
11-06 Arizona Department of Veterans’
Services—Veteran Home