Medicaid Program and State Children's Health Insurance Program (SCHIP); Payment Error Rate Measurement, 50490-50513 [07-4240]

Agencies

• Centers for Medicare & Medicaid Services
• DEPARTMENT OF HEALTH AND HUMAN SERVICES

[Federal Register Volume 72, Number 169 (Friday, August 31, 2007)]
[Rules and Regulations]
[Pages 50490-50513]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 07-4240]



[[Page 50489]]

-----------------------------------------------------------------------

Part III





Department of Health and Human Services





-----------------------------------------------------------------------



Centers for Medicare & Medicaid Services



-----------------------------------------------------------------------



42 CFR Parts 431 and 457



 Medicaid Program and State Children's Health Insurance Program 
(SCHIP); Payment Error Rate Measurement; Final Rule

Federal Register / Vol. 72, No. 169 / Friday, August 31, 2007 / Rules 
and Regulations

[[Page 50490]]


-----------------------------------------------------------------------

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Centers for Medicare & Medicaid Services

42 CFR Parts 431 and 457

[CMS-6026-F]
RIN 0938-AN77


Medicaid Program and State Children's Health Insurance Program 
(SCHIP); Payment Error Rate Measurement

AGENCY: Centers for Medicare & Medicaid Services (CMS), HHS.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: This final rule sets forth the State requirements to provide 
information to us for purposes of estimating improper payments in 
Medicaid and SCHIP. The Improper Payments Information Act of 2002 
(IPIA) requires heads of Federal agencies to estimate and report to the 
Congress annually these estimates of improper payments for the programs 
they oversee, and submit a report on actions the agency is taking to 
reduce erroneous payments.
    This final rule responds to the public comments on the August 28, 
2006 interim final rule (71 FR 51050) and sets forth State requirements 
for submitting claims and policies to the CMS Federal contractors for 
purposes of conducting fee-for-service and managed care reviews. This 
final rule also sets forth the State requirements for conducting 
eligibility reviews and estimating case and payment error rates due to 
errors in eligibility determinations.

DATES: Effective Date: These regulations are effective on October 1, 
2007.

FOR FURTHER INFORMATION CONTACT: Janet E. Reichert, (410) 786-4580.

SUPPLEMENTARY INFORMATION:

I. Background

A. The Improper Payments Information Act of 2002

    The Improper Payments Information Act of 2002 (IPIA), Pub. L. 107-
300, enacted on November 26, 2002, requires the heads of Federal 
agencies annually to review programs they oversee that are susceptible 
to significant erroneous payments, to estimate the amount of improper 
payments, to report those estimates to the Congress, and to submit a 
report on actions the agency is taking to reduce erroneous 
expenditures. The IPIA directed the Office of Management and Budget 
(OMB) to provide guidance on implementation. OMB defines ``significant 
erroneous payments'' as annual erroneous payments in the program 
exceeding both 2.5 percent of program payments and $10 million (OMB M-
03-13, May 21, 2003 and OMB M-06-23, August 10, 2006). For those 
programs with significant erroneous payments, Federal agencies must 
provide the estimated amount of improper payments and report on what 
actions the agency is taking to reduce them, including setting targets 
for future erroneous payment levels and a timeline by which the targets 
will be reached.
    According to the OMB directive, Federal agencies must include in 
the report to the President and Congress: (1) The estimate of the 
annual amount of erroneous payments; (2) a discussion of the causes of 
the errors and actions taken to correct those problems, including plans 
to increase agency accountability; (3) a discussion of the amount of 
actual erroneous payments the agency expects to recover; (4) 
limitations that prevent the agency from reducing the erroneous payment 
levels, that is, resources or legal barriers; and (5) a target for the 
program's future payment rate, if applicable.
    The Medicaid program and the State Children's Health Insurance 
Program (SCHIP) were identified by OMB as programs at risk for 
significant erroneous payments. OMB directed the Department of Health 
and Human Services (DHHS) to report the estimated error rates for the 
Medicaid and SCHIP programs each year for inclusion in the Performance 
and Accountability Report (PAR).
    Through the Payment Accuracy Measurement (PAM) and Payment Error 
Rate Measurement (PERM) pilot projects that CMS operated in Fiscal 
Years (FYs) 2002 through 2005, we developed a claims-based review 
methodology designed to estimate State-specific payment error rates for 
all adjudicated claims within 3 percent of the true population error 
rate with 95 percent confidence. An ``adjudicated claim'' is a claim 
for which either money was obligated to pay the claim (paid claims) or 
for which a decision was made to deny the claim (denied claims).

B. CMS Rulemaking

    Section 1102(a) of the Social Security Act (the Act) authorizes the 
Secretary to establish such rules and regulations as may be necessary 
for the efficient administration of the Medicaid and SCHIP programs. 
The Medicaid statute at section 1902(a)(6) of the Act and the SCHIP 
statute at section 2107(b)(1) of the Act require States to provide 
information that the Secretary finds necessary for the administration, 
evaluation, and verification of the States' program. Also, section 
1902(a)(27) of the Act (and 42 CFR 457.950) requires providers to 
submit information regarding payments and claims as requested by the 
Secretary, State agency, or both.
    Under the authority of these statutory provisions, we published a 
proposed rule on August 27, 2004 (69 FR 52620) to comply with the 
requirements of the IPIA and the OMB guidance. Based on the methodology 
developed in the pilot projects, the proposed rule set forth provisions 
for all States annually to estimate improper payments in their Medicaid 
and SCHIP programs and to report the State-specific error rates for 
purposes of our computing the national improper payment estimates for 
these programs. The intended effects of the proposed rule were to have 
States measure improper payments based on FFS, managed care, and 
eligibility reviews; to identify errors; to target corrective actions; 
to reduce the rate of improper payments; and to produce a corresponding 
increase in program savings at both the State and Federal levels.
    After extensive analysis of the issues related to having States 
measure improper payments in Medicaid and SCHIP, including public 
comments on the provisions in the proposed rule, we revised our 
approach. Our revised approach adopted the recommendation to engage 
Federal contractors to review State Medicaid and SCHIP fee-for-service 
(FFS) and managed care claims (we define the term ``claims'' to include 
both managed care capitation payments and FFS line items) and to 
calculate the State-specific and national error rates for Medicaid and 
SCHIP. States will calculate the State-specific eligibility error 
rates. Based on these rates, the Federal contractor will calculate the 
national eligibility error rate for each program. We also adopted the 
recommendation to sample a subset of States each year rather than to 
measure every State every year. We adopted these recommendations 
primarily in response to commenters' concerns with the cost and burden 
to implement the regulatory provisions at the State level that the 
proposed rule would have imposed on States.
    Since our revised approach departed significantly from the approach 
in the proposed rule, we published an interim final rule with comment 
period on October 5, 2005 (70 FR 58260). The October 5, 2005 interim 
final rule with comment period responded to the public comments on the 
proposed rule, and informed the public of our national contracting 
strategy and of our plan to measure improper payments in a subset

[[Page 50491]]

of States. Our State selection will ensure that a State will be 
measured once, and only once, every 3 years for each program. For each 
fiscal year, we stated that we expected to measure up to 18 States. We 
also stated that we would use a rotational approach to review the 
States' Medicaid programs. The rotation allows States to plan for the 
reviews because States know in advance in which year they will be 
measured. At the end of the first 3-year cycle, the rotation will 
repeat so that the FY 2006 States will be reviewed again in FY 2009; 
the FY 2007 States will be reviewed again in FY 2010; and the FY 2008 
States will be reviewed again in FY 2011. The rotation will continue in 
this manner for future years.
    In determining the Medicaid State selection, we grouped all States 
into three equal strata of small, medium, and large, based on the 
States' most recently available FFS annual expenditure data. We 
randomly selected up to six States from each stratum each year, until 
we selected all States for the first cycle of FY 2006 through FY 2008. 
We announced the Medicaid State selection rotation in the October 5, 
2005 interim final rule and also through a State Health Official Letter 
released to all States on November 18, 2005.
    In the October 5, 2005 interim final rule, we stated that it was 
still possible that States sampled for review would be required to 
conduct eligibility reviews as described in the proposed rule. We also 
announced our intentions to establish an eligibility workgroup to make 
recommendations on the best approach for reviewing Medicaid and SCHIP 
eligibility within the confines of current statute, with minimal impact 
on States and additional discretionary funding. We convened an 
eligibility workgroup comprised of DHHS (including CMS and, in an 
advisory capacity, the Office of the Inspector General (OIG)), OMB, and 
representatives from two States. We determined that States should 
conduct the eligibility measurement and developed an eligibility 
measurement methodology based on the workgroup's consideration of 
public comments, the examination of various approaches proposed in such 
comments, and the suggestions of the panel members.
    The October 5, 2005 interim final rule also set forth the types of 
information that States would submit to the Federal contractors for the 
purpose of estimating Medicaid and SCHIP FFS improper payments and 
invited further comments on methods for estimating eligibility and 
managed care improper payments. We received very few comments regarding 
managed care and a number of comments regarding eligibility.
    Based on the public comments and recommendations from the 
eligibility workgroup, we published a second interim final rule on 
August 28, 2006 (71 FR 51050), which set forth the methodology for 
measuring improper payments in Medicaid and SCHIP FFS, managed care, 
and eligibility in 17 States and invited further public comments on the 
eligibility measurement.

C. IPIA Compliance

    We expect to be fully compliant with IPIA requirements by the year 
2008. We measured Medicaid FFS improper payments in FY 2006 and plan to 
have all components (FFS, managed care, and eligibility) of Medicaid 
and SCHIP measured in FY 2007 for reporting in the FY 2008 Performance 
and Accountability Report (PAR).
    These measurements in 17 States each year will produce State-
specific component error rates as well as composite program error rates 
for the State's Medicaid and SCHIP programs. From the State-specific 
error rates, we will calculate national error rates for each of the 
components and for the Medicaid and SCHIP programs.
    We expect State corrective actions to address the causes of error 
in each of the program components. As a result, we expect States will 
reduce their program error rates over the course of each measurement 
cycle which, in turn, should reduce the national error rates.

II. Provisions of the August 28, 2006 (Second) Interim Final Rule

    We published a second interim final rule with comment period on 
August 28, 2006 that responded to comments on the October 5, 2005 
initial interim final rule with comment period. In the August 28, 2006 
interim final rule, we reiterated our national contracting strategy to 
estimate improper payments in both Medicaid and SCHIP fee-for-service 
and managed care claims and set forth the State requirements for 
estimating improper payments due to errors in Medicaid and SCHIP 
eligibility determinations. We also announced that a State's Medicaid 
and SCHIP programs would be reviewed in the same year.

A. Selecting SCHIP States for Review

    After the October 2005 Medicaid State selection, we decided on the 
SCHIP State selection for the PERM measurement beginning with FY 2007. 
We determined that SCHIP could be measured in the same States selected 
for Medicaid review each fiscal year with a high probability that the 
SCHIP error rate would meet OMB requirements for confidence and 
precision levels.
    We believe that paralleling the SCHIP and Medicaid measurements 
will minimize administrative complexities for both CMS and the States. 
Measuring both programs at the same time may further reduce the State 
cost and burden because States are able to plan activities for both 
measurements and may gain efficiencies by combining staff and resources 
for the reviews.
    We announced in the August 28, 2006 interim final rule our decision 
to measure Medicaid and SCHIP in a State at the same time. We also sent 
a State Health Official Letter to all States regarding the SCHIP State 
selection on August 30, 2006. As with Medicaid, we stated that we 
expected to measure improper payments in all components of SCHIP in FY 
2007 and beyond. The selection of States for the first PERM cycle of FY 
2006 through FY 2008 is listed below. Note that, for States measured 
for Medicaid FFS in FY 2006, all three components of Medicaid and SCHIP 
will be measured in FY 2009.

                   Medicaid and SCHIP State Selection
------------------------------------------------------------------------
 
------------------------------------------------------------------------
FY 2006.....................  Pennsylvania, Ohio, Illinois, Michigan,
                               Missouri, Minnesota, Arkansas,
                               Connecticut, New Mexico, Virginia,
                               Wisconsin, Oklahoma, North Dakota,
                               Wyoming, Kansas, Idaho, Delaware.
FY 2007.....................  North Carolina, Georgia, California,
                               Massachusetts, Tennessee, New Jersey,
                               Kentucky, West Virginia, Maryland,
                               Alabama, South Carolina, Colorado, Utah,
                               Vermont, Nebraska, New Hampshire, Rhode
                               Island.
FY 2008.....................  New York, Florida, Texas, Louisiana,
                               Indiana, Mississippi, Iowa, Maine,
                               Oregon, Arizona, Washington, District of
                               Columbia, Alaska, Hawaii, Montana, South
                               Dakota, Nevada.
------------------------------------------------------------------------


[[Page 50492]]

B. PERM Measurement Cycle

    We stated in the August 28, 2006 interim final rule that the 
process for measuring improper payments, called the ``production 
cycle,'' under the national contracting strategy would take 
approximately 23 months per cycle. Using FY 2006 as an example, we 
provided the following table as an approximate overview of the PERM 
process. It is important to note that the process is fluid, so 
timeframes may fluctuate slightly depending on such factors as the 
complexities of the reviews.

              Example of the PERM Production Cycle: FY 2006
                  [Note: only illustrates Medicaid FFS]
------------------------------------------------------------------------
             Timeframe                              Event
------------------------------------------------------------------------
December 1, 2005..................   States submit medical
                                     policies in effect for the review
                                     period to the DDC.
January 15, 2006..................   States submit 1st quarter
                                     FY 2006 (October-December 2005)
                                     adjudicated claims to the SC.
February 1, 2006..................   State submits 1st quarter
                                     FFS policy updates to the DDC.
April 15, 2006....................   States submit 2nd quarter
                                     FY 2006 (January-March 2006)
                                     adjudicated claims to the SC.
May 1, 2006.......................   States submit 2nd quarter
                                     policy updates to the DDC.
July 15, 2006.....................   States submit 3rd quarter
                                     FY 2006 (April-June 2006)
                                     adjudicated claims to the SC.
August 1, 2006....................   States submit 3rd quarter
                                     policy updates to the DDC.
October 15, 2006..................   States submit 4th quarter
                                     FY 2006 (July-September 2006)
                                     adjudicated claims to the SC.
November 1, 2006..................   States submit 4th quarter
                                     policy updates to the DDC.
Throughout PERM process...........   States identify and resolve
                                     differences in review findings with
                                     the RC.
------------------------------------------------------------------------

C. Use of Federal Contractors to Review FFS and Managed Care Claims

    In the August 28, 2006 interim final rule, we reiterated that, 
under the national contracting strategy, we would use Federal 
contractors to measure Medicaid and SCHIP FFS and managed care improper 
payments. We believe the use of more than one CMS Federal contractor 
allows for the award of contracts in areas of specialization and 
expertise, minimizes potential problems with the error rate measurement 
process if one contractor experiences operational difficulties, and 
provides us with optimum oversight. However, we may revise our use of 
multiple contractors in the future if warranted by our experience as 
the program matures, for example, if we can gain efficiencies. For FYs 
2006 and 2007, we awarded three contracts: (1) A statistical analysis 
contract; (2) a documentation/database contract; and (3) a review 
contract.
    The statistical contractor (SC) collects adjudicated claims data, 
determines the sample size, draws the sample, and calculates the State 
and national error rates. The documentation/database contractor (DDC) 
standardizes State data, collects and stores State medical and other 
related policies, and requests the medical records from providers for 
the FFS medical reviews. The review contractor (RC) conducts the 
medical and data processing reviews on the States' FFS and managed care 
claims.
    In the August 28, 2006 interim final rule, we indicated that the 
States' responsibilities to support the improper payments measurement 
for both Medicaid and SCHIP would include submission of information on 
managed care. We stated that the States selected for review would 
submit to the SC the following information for Medicaid and SCHIP:
     All adjudicated FFS and managed care claims information 
from the review year on a quarterly basis, with FFS claims stratified 
into seven strata by service type and one additional stratum for denied 
claims;
     Information on claims that were selected as part of the 
sample, but which changed in substance after selection (for example, 
successful provider appeals); and
     Adjustments made within 60 days after the adjudication 
dates for the original claims or line items, with sufficient 
information to indicate the nature of the adjustments and to match the 
adjustments to the original claims or line items.
    We required States to provide stratified FFS claims data because we 
believed that stratifying the claims by service type would improve the 
efficiency of the sampling methodology by distributing the claims in 
the sample in proportion to the dollar share in the universe. 
Stratification allows services with a larger dollar share to compose a 
larger share of the sample and reduces the variance in the sample. 
Stratifying the claims also allows for smaller sample sizes and for the 
identification of errors in specific service types so that States would 
have information that could be helpful to target causes of errors.
    Based on the annual expenditure data, the SC would determine the 
State's sample size and, for FFS claims, the sample size for each of 
the eight total strata. These strata were established during the pilot 
projects based on the total share of dollars. States had already 
grouped their claims similarly in their Medicaid Management Information 
System (MMIS); therefore, we believed that the stratification of claims 
for submission would not be burdensome to States.
    We established the following strata: (1) Hospital services; (2) 
long term care services; (3) other independent practitioners and 
clinics; (4) prescription drugs; (5) home and community based services; 
(6) other services and supplies (for example, durable medical 
equipment, clinical lab tests, and x-rays); (7) primary care case 
management; and (8) denied claims.
    From the State's quarterly adjudicated claims data, the SC would 
randomly select a sample of FFS and managed care claims each quarter. 
Each selected FFS claim would be subjected to a medical and data 
processing review. Managed care claims would not be stratified or 
subjected to medical reviews because the payments made to a managed 
care plan are based on a set fee from a predetermined capitation 
agreement, rather than for the specific service(s) provided. We 
expected that the sample size would be 1,000 FFS claims and 500 managed 
care claims per State per program in order to achieve a 3 percent 
precision level at the 95 percent confidence level (based on a range 
estimated during the PAM/PERM pilots).
    For review of the sampled claims, States would provide the DDC the 
following information for Medicaid and SCHIP:
     All medical and other related policies in effect for the 
review year and any quarterly policy updates;
     Current managed care contracts, rate information, and any 
quarterly updates to contracts and rates for the review year for SCHIP 
and, as requested, for Medicaid; and

[[Page 50493]]

     Upon request from the contractor, provider contact 
information that has been verified by the State as current.
    States selected for review also would provide the RC the following 
information for Medicaid and SCHIP:
     Systems manuals for data processing reviews. (If a State's 
medical and data processing policies are intertwined, the State may 
send the policies to the DDC. The DDC would then identify the data 
processing policies so the RC could access them through the DDC.)
     Repricing information, as requested by the RC, for claims 
that the RC determined to be improperly paid. The RC would request that 
States submit the price that should have been paid so that, for claims 
that were found to be in error, the RC would be able to determine the 
amount of the improper payment.
    The August 28, 2006 interim final rule also set forth a difference 
resolution process whereby States would be provided disposition reports 
listing the contractor's review finding on each claim. Based on these 
reports, States would be able to dispute error findings.
    When the reviews were completed, the SC would estimate the State-
specific error rates for the FFS and managed care components of the 
Medicaid and SCHIP programs. States (using the eligibility methodology 
set forth in the August 28, 2006 interim final rule to conduct 
eligibility reviews beginning in FY 2007) would calculate and report 
the State-specific eligibility error rates to us. These measurements 
also will produce component error rates for the State's Medicaid and 
SCHIP programs. From the State-specific error rates, we will calculate 
national error rates for each of the components and for the Medicaid 
and SCHIP programs.
    Once the State-specific and national error rates were estimated, 
the States would develop and send to us corrective action reports 
describing corrective actions that the States would implement to 
address the major causes of improper payments. The States would review 
their error rates, determine root causes of error-prone areas, and 
develop corrective actions to address the major error causes for 
purposes of reducing the payment error rates. States selected for 
review would provide us with the following information for Medicaid and 
SCHIP:
     A corrective action report for purposes of reducing the 
State's payment error rates in the FFS, managed care, and eligibility 
components of the program; and
     Other information that the Secretary determined necessary 
for, among other purposes, estimating improper payments and determining 
error rates in Medicaid and SCHIP.
    We stated that we would request information we found during the 
course of measuring each program that would improve the process, 
produce more accurate error rates, or reduce the cost and burden on 
either or both the State and Federal governments. Similarly, we stated 
that, if we determined that we were collecting specific information 
that did not add value to the error rate measurement or was not 
productive to collect, we would discontinue that collection.

D. Eligibility Measurement

    In the August 28, 2006 interim final rule, we set forth the 
eligibility measurement methodology developed through the eligibility 
workgroup and through our consideration of public comments submitted in 
response to the October 5, 2005 initial interim final rule. The 
eligibility measurement methodology is summarized below:
     A State would review program eligibility in the year it 
was scheduled for review for FFS and managed care improper payments. 
The eligibility reviews would be conducted by a State agency that was 
functionally and physically independent of the State agency making the 
program policy and eligibility determinations.
     The Medicaid and SCHIP eligibility sample universes would 
consist of both active cases (individuals enrolled in the program) and 
negative cases (individuals denied or terminated from the program).
     Medicaid and SCHIP cases in the active universe would be 
stratified into three strata: (1) Applications, (2) redeterminations, 
and (3) all other cases. Negative case action samples would not be 
stratified in either program.
     A State would calculate its eligibility error rates for 
active cases (including undetermined cases) and negative cases.
     States would submit the following to CMS:

--A sampling plan for approval (which would be submitted 60 days before 
the beginning of the fiscal year selected for review);
--A monthly sample selection list that identified the cases selected 
for review (to be submitted each month and before commencing the 
reviews);
--Detailed findings on the cases reviewed;
--Summary findings on the cases reviewed; and
--State-specific case and payment error rates for active cases, case 
error rates for negative cases, the number and amount of undetermined 
cases, and the total amount of payment from all undetermined cases in 
the active case sample, to be submitted by July 1 after the end of the 
fiscal year under review.

    We invited further comment on this methodology for measuring 
improper payments due to errors in eligibility determinations.

III. Analysis of and Responses to the Public Comments on the August 28, 
2006 Interim Final Rule

    We received a total of 33 comments: 28 from State agencies, 3 from 
consumer advocacy and other groups and 2 from individuals. These 
commenters reiterated some of the comments from the proposed rule to 
which we responded in the October 5, 2005 and August 28, 2006 interim 
final rules. Although we are not required to respond to these comments 
again, we are summarizing the comments in this final rule and providing 
our responses for the convenience of the reader. Below are the comments 
on the August 28, 2006 interim final rule and our responses.
    Most comments responded to our invitation for further comment on 
the PERM eligibility measurement process. Commenters also indicated 
that, although the August 28, 2006 interim final rule significantly 
reduced the burden on the States by using a Federal contracting 
strategy and limiting State selection to once every 3 years, they 
believed that the August 28, 2006 interim final rule still placed an 
undue technical and financial burden on the States to assist the 
Federal contractors.

A. Purpose, Basis, and Scope

1. Payment Error Rates
    Comment: Several commenters asserted that a State error rate is not 
required by IPIA and funds are wasted in establishing a payment error 
rate. The commenters also maintained that State audits could identify 
improper payments. The commenters stated that a national sampling 
framework should be used to measure a national error rate, and that CMS 
should abandon the proposed State-level error rate in favor of a 
national error rate and sampling plan.
    Response: As we observed in the October 5, 2005 and August 28, 2006 
interim final rules, the IPIA requires the Secretary to estimate the 
amount of improper payments in programs and activities that are 
susceptible to significant improper payments and report those estimates 
to the Congress. OMB has identified Medicaid and SCHIP as programs at 
risk for significant

[[Page 50494]]

improper payments. Because States administer these programs and because 
there is wide variation in States' coverage, eligibility, benefit, and 
reimbursement policies for these programs, we must rely on State-
specific information to develop State-level estimates as the basis for 
a national program error rate.
    In addition, even though State audits may identify improper 
payments, we could not be confident that States' audit procedures would 
be similar and would be consistently applied nationwide or would 
produce statistically reliable information on which a national rate 
could be based. Finally, we have stated that the PERM program is 
intended to fulfill the requirements of the IPIA; it is not intended to 
supplant, enhance, or change other program integrity activities in 
which the States are currently engaged.
    Comment: One commenter suggested that the national error rate be 
computed using State error rates that are weighted against dollar 
volume in other States to ensure that each State's contribution to the 
error rate is clear, balanced, and consistently calculated at all 
levels of data analysis.
    Response: The national error rate is calculated as the outcome of a 
two-stage sampling process. States were placed into one of three 
strata. These strata consist of the large, medium, and small States as 
measured by Medicaid expenditures. For each of the three rotations, 17 
States were randomly selected from three strata. Beginning in FY 2007, 
for the States sampled in each year, claims and payments are sampled 
for Medicaid and SCHIP fee-for-service and managed care. Sufficient 
numbers of claims and payments are sampled to estimate an error rate 
for the State at a precision level of plus or minus 3 percentage points 
with 95 percent confidence. Then, within each of the three strata, an 
error rate is calculated to represent the error rate of that stratum. 
Finally, a national error rate is calculated by computing the error 
rates across the three strata, where each stratum's rate is weighted by 
the share of expenditures for that program represented by its strata. 
The variance in this estimate is calculated by taking into account: (1) 
The variance of the error rate of the individual States in the sample, 
and (2) the variance in the original sample of States from the three 
strata. The error rate is based on the total error, not the State or 
Federal share.
    Comment: One commenter suggested that States should be allowed to 
calculate error rates based on either the difference method or ratio 
method.
    Response: Our statistical contractor will calculate the State-
specific error rates for FFS and managed care. In general, the ratio 
method of estimating the error rate is formed using data from the 
sample. From the sample, the dollar value of claims or payments in 
error enters the numerator, while the dollar value of payments (both 
those made in error and those that are valid) enters into the 
denominator. This ratio is the error rate.
    In general, the ``difference'' estimator is calculated as follows. 
The dollar value of each error (the difference between what should have 
been paid and what was paid) in the sample is added, with weights equal 
to the inverse of the sampling frequency for the respective claim or 
line item. This provides an estimate of the total dollars in error for 
the universe or population for which the inference is made. This 
becomes the numerator of the error rate. The denominator of the error 
rate is actual payments made for the universe or population. The 
denominator is non-stochastic, that is, non-random. This ratio, then, 
provides an estimate of the error rate.
    Because the actual payments made by the State for the universe or 
population may not be available when we calculate the error rate, we 
plan to use the ratio estimator.
    Comment: A commenter observed that, in the August 28, 2006 interim 
final rule, we responded to a comment regarding the likelihood of 
achieving a national error rate by aggregating error rates from all the 
States' programs with their inherent variations. We stated that, ``(b)y 
drawing a stratified random sample of States and then reviewing a 
random sample of claims within each of those States (using each State's 
program policies), we are able to obtain an estimate of the national 
error rate without having to conduct reviews on all claims. This 
methodology will produce the estimate and the precision level of the 
estimated national error rate, within the parameters set by OMB.'' The 
commenter asserted this logic is circular and stated that more 
information is needed to explain how this process would work.
    Response: The process is based on sampling. By sampling, one can 
obtain an estimate of a population parameter, such as the mean dollar 
value of a Medicaid claim for a State, without having to examine every 
claim in that State's universe. The larger the sample size, the more 
precise the estimate of the mean value will be. For most populations, 
one can typically obtain a very precise estimate of the population 
parameters, such as the mean, by sampling far fewer than the entire 
population or universe. Based on the outcome of the sample, one can 
make an inference regarding the values of the true population mean, for 
example, and a statement of the probability or likelihood that a small 
range around the sample estimate captures the population's true mean.
    The national error rate is calculated as the outcome of a two-stage 
sampling process. First, States are sampled. Then, claims are sampled 
within the State.
    States were placed into one of three strata. These strata consist 
of the large, medium, and small States as measured by Medicaid 
expenditures. For each year, a total of 17 States were randomly 
selected from the three strata. For States sampled in each strata, 
claims and payments are sampled for Medicaid and SCHIP fee-for-service 
and managed care. A sufficient number of claims and payments are 
sampled to estimate an error rate at a precision level of plus or minus 
3 percentage points with 95 percent confidence for that State. Then, 
within each of the three strata, an error rate is calculated based on 
the States sampled in that stratum. Finally, a national error rate is 
calculated by estimating the error rate for the population of all 
States as a weighted average of the error rates within each stratum. 
The variance in this estimate is calculated by taking into account the 
variance of the error rate of the individual States in the sample and 
the variance in the original sample of States from the three strata.
    Comment: A commenter would like to know the operational benefit of 
a national error rate to the States if they will be measured against 
their individual rates rather than a national average.
    Response: The Improper Payments Information Act of 2002 (IPIA) 
requires CMS to estimate and report to the Congress annual estimates of 
improper payments. The national error rate for SCHIP and Medicaid will 
be reported to the Congress as required by law. States will use their 
State-specific error rates to implement corrective action plans. We 
believe that these plans will ultimately reduce the national error 
rate.
    Comment: A commenter asked what assurance States would have that 
comparisons among States would not be made when the error rates were 
reported. Because of the wide variation in States' Medicaid and SCHIP 
programs, this assurance is needed in order to reassure States that 
unwarranted comparisons are not being made.

[[Page 50495]]

    Response: We agree that care should be taken in comparing the State 
error rates due to variation in State programs.
    Comment: A commenter requested that CMS develop methods to 
communicate with States regarding their responsibilities, timelines, 
and completion expectations.
    Response: We have communicated with States through kick-off calls 
and one-on-one calls with each State involved in each year's 
measurement. In addition, we post all instructions, letters and 
questions and answers on our CMS PERM Web site at https://
www.cms.hhs.gov/PERM for all States to review.
    Comment: A commenter stated that, since PERM is measured in a 3-
year cycle, the ``national average'' error rate cannot be compared 
year-to-year.
    Response: We believe there are several approaches to assess the 
improvement in the reduction of improper payments year-to-year and over 
the years.
    Comment: Two commenters believed that State program integrity 
efforts are in jeopardy because claims from providers under active 
fraud investigation are included in the universe. The commenters 
believed that (1) The error rate will be inflated because fraudulent 
and abusive providers are not likely to respond to requests for medical 
records; (2) providers can create, alter, or destroy documentation and 
evidence when they are alerted that their claims are investigated; and 
(3) false, fraudulent, and abusive claims can only be identified by 
interviewing recipients and reviewing medical records.
    Response: We do not intend to jeopardize States' provider fraud 
investigations based on our review of FFS and managed care claims. 
Therefore, if a FFS or managed care claim sampled under PERM is part of 
a fraud investigation and the State notifies the statistical contractor 
of this fact, the claim will not be subject to review under PERM. 
However, we will cite the claim as an error. We believe the State, in 
this instance, also believes the claim is in error since the State is 
investigating the provider for fraud. For purposes of the eligibility 
review, which is conducted on individual beneficiary cases rather than 
claims, cases under beneficiary fraud investigation are excluded from 
review.
    Comment: A commenter asked for clarification on whether the IPIA is 
intended to root out provider fraud or challenge program enrollment 
decisions. The commenter stated that those functions are under the 
purview of other Federal and State initiatives.
    Response: We agree that the IPIA is not intended to root out these 
problems. The IPIA is intended to identify improper payments, and 
provider fraud may be discovered during the course of the measurement. 
In addition, erroneous Medicaid and SCHIP program enrollment decisions 
may be discovered during the eligibility reviews. The discovery of 
these problems would be addressed by the State through corrective 
actions.
    Comment: A commenter indicated that the rule does not explain if 
extrapolations will be conducted and if error rates will be reported 
based on claims, dollar amount, or both.
    Response: The method for estimating error rates is based on 
sampling from the population or universe. From the sample, inferences 
(or extrapolations) are drawn regarding specific population or universe 
values, such as the error rate for the population. The active case 
eligibility error rates will be dollar-weighted error rates. The 
dollars assigned to the case will be those associated with the claims 
that are collected for the recipient. The sample sizes for the active 
cases were constructed to achieve an estimate of the State's dollar-
weighted error at a precision level of plus or minus 3 percentage 
points with 95 percent confidence. The State level active case 
eligibility error rates will be a component of a national active case 
eligibility error rate. A simple binomial error rate (valid/invalid) 
will be calculated for the active case error rate, and a binomial 
(valid/invalid) error rate will be calculated for negative cases.
    The Medicaid and SCHIP error rates for both fee-for-service and 
managed care will be calculated and reported based on the dollar value 
of the line items or payments sampled. The sample sizes were 
constructed to achieve a precision level for each of the programs 
(Medicaid and SCHIP fee-for-service and managed care) of plus or minus 
3 percentage points with 95 percent confidence. The State level error 
rates will also be used to estimate national error rates for these 
programs, which are expected to achieve a precision level of plus or 
minus 2.5 percentage points with 90 percent confidence.
    To summarize, the methodology is to sample from the population or 
universe, and then use the sample results to infer or extrapolate the 
error rate for the population.
2. State Selection
    Comment: A commenter stated that States were led to believe that 
each program would be measured on an alternating or rotational basis. 
By measuring Medicaid and SCHIP in the same year, the commenter 
believes that CMS has unilaterally increased the State's cost and 
burden by 100 percent. According to the commenter, this decision is 
contrary to the supporting statement issued with the initial request to 
gain OMB approval (71 FR 30409) published May 26, 2006.
    Response: We believe that State cost and burden could actually be 
reduced by measuring both programs in the same year. States would have 
to measure errors in both programs at some point. By evaluating them 
simultaneously, we believe efficiencies will be gained that may lower 
costs and burden. We stated in both published interim final rules that 
we would rotate the States, not the programs. We reiterate, in this 
final rule, that each State will be measured on a rotational basis.
    Comment: A commenter stated that the proposed random selection of 
States to be reviewed under the PERM program makes it difficult to 
predict the resources needed for PERM-related activities. If not 
forthcoming, States could be held responsible for time delays in the 
program.
    Response: In the August 28, 2006 interim final rule, we stated that 
we will use a rotational approach to review the States in Medicaid and 
SCHIP. We released instructions explaining the selection of the States 
to be reviewed under the PERM program through an October 10, 2006 State 
Health Official letter. This information was also posted on the CMS 
PERM Web site at https://www.cms.hhs.gov/PERM. Further, we stated that 
we believe that the rotation will allow States to plan for the reviews 
because States will know in advance in which year they will be 
measured.
3. Use of National Contractor
    Comment: A commenter stated that Generally Accepted Government 
Auditing Standards (GAGAS) require States to review and comment on 
contractor-generated PERM working papers and findings for quality 
control purposes. The commenter asserted that the contractor's findings 
should not be deemed final or actionable until this review is complete. 
In addition, the commenter stated that the cost of this review must be 
included in the rules, which, according to the commenter, does not 
appear to be the case.
    Response: The PERM program does not require States to use GAGAS. 
GAGAS is issued by the Comptroller of the United States as auditing 
standards for governmental audits. The PERM program is not an audit and 
as such, GAGAS would not be applicable. However, under PERM, States 
have the

[[Page 50496]]

opportunity through the difference resolution process to review error 
findings. States also have the opportunity to further dispute error 
findings by appealing to CMS.
    Comment: A commenter applauded the use of national contractors but 
did not believe the contractors have the required knowledge to complete 
the reviews under CMS' current schedule. The commenter believed 
additional time is needed for the transfer of knowledge from State to 
contractor.
    Response: The contractors will work closely with the States during 
the measurement process to ensure that program knowledge is 
transferred. We believe this will help mitigate delays in the process 
that might be encountered otherwise.
    Comment: A commenter asked how many days after the quarter ends 
would State information have to be submitted to the statistical 
contractor. The commenter stated that no details were provided on page 
51053 of the Federal Register publication of the August 28, 2006 
interim final rule.
    Response: Our statistical contractor's instructions request that 
State information be submitted to the statistical contractor no later 
than 15 days after the quarter ends.
    Comment: A commenter asked CMS to further clarify the format in 
which States will be required to submit data for PERM compliance 
purposes and whether the data would need to be coded.
    Response: The operational details are contained in the instructions 
that the statistical contractor sends to the States being measured at 
the beginning of each quarter.
    Comment: A commenter stated that the delay in collecting provider 
documentation does not allow enough time for a State to respond to any 
findings or perceived errors. The commenter does not believe that 
hiring three contractors is effective in measuring error rates.
    Response: We believe that having three contractors is effective 
because the program is not jeopardized or substantially delayed if one 
contractor experiences problems; the other contractors could continue 
their respective aspects of the measurement. We agree that the 90-day 
timeframe to collect medical records from providers may not allow 
States adequate time to resolve errors with the RC through the 
difference resolution process. In order to expedite the difference 
resolution process within the overall timeframe for calculating annual 
error rates under PERM, and provide States with adequate time to 
respond to our contractor's proposed findings, we will issue guidance 
instructing our national contractors to request that providers, in 
compliance with our regulations at 42 CFR 431.107(b)(2), 431.970, and 
457.720, submit medical records no later than 60 days after issuance of 
the contractor's letter requesting such records. This will provide 
additional time for the State and contractor to analyze and resolve 
discrepancies.
4. State Input into the Program
    Comment: One commenter disagreed with CMS' statement that States 
have been active participants in the PERM regulatory process. The 
commenter stated that CMS has not provided an acceptable forum for 
State participation in the development of PERM regulation, and that 
only two States were involved in meetings with CMS during the 
development of the regulation. In addition, the commenter indicated 
that CMS has not been present on three all-State calls regarding PERM 
regulation, and that when CMS is present on calls, CMS does not provide 
substantive responses to questions and points of clarification from the 
States. The commenter concluded that States cannot make reasonable 
comments and suggestions when CMS does not provide States with 
sufficient information.
    Response: The two States participated in the eligibility workgroup; 
they did not participate in developing the entire PERM regulation. 
Consistent with the rulemaking process, we have provided a vehicle by 
which we review all timely public comments submitted to us. Through 
this process, we have received valuable assistance in developing an 
error rate measurement procedure that we believe is both sensitive to 
the burdens that States must bear in meeting their responsibilities, as 
well as one that allows us to uphold the duties that we must carry out 
to be in compliance with the IPIA.

B. Methodology

    Comment: Commenters stated that CMS should provide a detailed 
timeline for the PERM sampling year for claim and eligibility reviews, 
so that States would understand the schedule and deadlines. They 
indicated that this timeline should identify all three contractor 
activities and expected State responsibilities (for example, claim 
delivery and sampling schedule dates and required State documentation 
due dates needed by contractors to comply with CMS contract deadlines). 
In addition, the commenters noted that States have suggested that, for 
each PERM State being reviewed, the contractors should prepare monthly 
project planning documents to CMS and the States that would explain 
delays, barriers, or other issues that have arisen and the contractor's 
plans to resolve any problem areas.
    Response: We provided an overall timeline of the measurement 
process in the August 28, 2006 interim final rule (using the FY 2006 
Medicaid FFS measurement as an example) to identify when States should 
submit needed information. We have included the timeline again in this 
final rule (see ``Example of the PERM Production Cycle: FY 2006'' 
illustration) for the reader's convenience. In addition, we have held 
kick-off calls, State-specific calls, component review calls, and 
provided instructions to States selected for the FY 2006 and FY 2007 
measurements, so States would understand the schedule and deadlines for 
the FFS and managed care claims data submission. We intend to provide 
the same guidance to States selected for the FY 2008 and FY 2009 
measurements. The timeline for the eligibility measurement is attached 
to the eligibility instructions, which can be found along with the 
claims submission instructions, on the CMS PERM Web site at https://
www.cms.hhs.gov/PERM.
Sampling
1. Exclusions From the Claims Universe
Denied Claims
    Comment: Two commenters suggested that CMS remove denied claims as 
a review stratum. The commenters stated that there is an increased 
burden on States to produce a list of adjudicated denied claims and 
track re-billings of denied claims. The commenters also noted that 
there is difficulty in determining the sample size based on dollar 
value when the value of the denied claim is zero. The commenters 
recommended convening a workgroup to determine a methodology to measure 
errors in denied claims.
    Response: Denied claims could be underpayments, and IPIA requires 
the inclusion of underpayments in our measurement. We believe it is as 
important to know when claims and eligibility have been wrongfully 
denied as when they have been wrongfully paid and approved. 
Furthermore, the sample size is determined by our statistical 
contractor, not the States. Finally, the methodology to measure errors 
in denied claims was developed by CMS and States during PAM/PERM 
pilots. Therefore, we are not adopting the suggestion to convene a 
workgroup to revisit this matter.

[[Page 50497]]

2. Sampling Issues
    Comment: A commenter noted that the PERM stratification 
requirements are complex and would likely pose a challenge for its 
systems.
    Response: We agree that stratifying Medicaid FFS claims has posed 
challenges for States. Many States measured in FY 2006 had difficulties 
stratifying the claims. Therefore, we are revising the requirement at 
Sec.  431.970(a)(1) to remove the stratification of Medicaid and SCHIP 
FFS claims by service requirement. This approach will further reduce 
State burden since States would need only to submit the universe data. 
We believe we can achieve greater sampling efficiency by stratifying 
the FFS claims by dollar value rather than by service. The Federal 
contractor will stratify the claims by dollar value.
    Comment: A commenter stated that CMS did not provide a rationale 
for the following statement in the August 28, 2006 interim final rule: 
``We did not adopt the recommendation to select a nationwide sample 
because we believed that it was not the best overall method to meet the 
requirements of the IPIA and OMB guidance. There is no national 
sampling framework for SCHIP claims * * *'' The commenter maintained 
that the absence of a national sample framework for SCHIP does not mean 
that one could not or should not exist.
    Response: We do not believe a national sample is the best method to 
achieve IPIA compliance. The Medicaid and SCHIP programs are State-
administered, and as such, we think it is necessary for States to 
participate in part of the measurement process. We considered the 
suggestions made by commenters on the past interim final rules and 
determined that we would not adopt this recommendation.
    Comment: A commenter asked whether the universe of claims includes 
pharmacy, mental health, and substance abuse claims.
    Response: Yes, pharmacy, mental health, and substance abuse claims 
are included in the universe of claims.
    Comment: Since the annual sample size is 1,000 FFS claims per State 
per program, a commenter stated that the State's SCHIP program will 
likely be disproportionately oversampled, since its State represents 
only approximately 10 percent of the total United States population.
    Response: From a sampling perspective, there is generally no 
difference between a small and large population. Specifically, a 
property of sampling is that, once the population size exceeds about 
10,000, the population can be treated as if it were an infinite 
population. In other words, statistically speaking, beyond a universe 
of about 10,000, population differences do not have a significant 
effect on sample size.
    Comment: A commenter asked CMS to clarify each sample size and 
methodology for each area of the PERM project. The commenter stated 
that in all correspondence released by CMS to the States, the sample 
sizes and methodologies have varied, which has made it difficult for 
States to determine what is expected from them.
    Response: PERM measures three components in Medicaid and three 
components in SCHIP: FFS, managed care, and eligibility. For FY 2006, 
the FFS sample size is 1,000 claims annually per program. These claims 
are subject to data processing and medical reviews by our contractor. 
For FY 2006, the managed care sample is 500 claims annually per 
program. These claims are subject to data processing review only by our 
contractor. For FY 2006, the eligibility sample size is 504 active 
cases and 204 negative cases (not claims). Reviews to verify 
eligibility are done by the States. Future sample sizes are subject to 
change as necessary depending on such factors as lessons learned or 
other situations impacting the timely and accurate error rate 
measurement.
    Comment: Commenters asked when FY 2007 States could expect to 
receive additional information regarding the data elements that would 
be required for data submission.
    Response: The statistical contractor sends instructions out to each 
State 45 days before the beginning of each fiscal year.
3. Medical Records Collection
    Comment: A commenter asked if it was the State's responsibility to 
pursue information identifying which providers have not submitted 
requested medical records and whether the documentation/database 
contractor would provide this information to the State.
    Response: The documentation/database contractor will request the 
medical records directly from providers for the FFS medical reviews. 
The contractor will follow-up with providers who have not submitted 
medical records. The contractor will notify the State of providers who 
have not submitted medical records. The State can opt to follow-up with 
these providers.
    Comment: A commenter stated that, when the documentation/database 
contractor receives the medical records from the provider, it is 
imperative that the contractor immediately review the records for 
completeness and appropriateness of documentation. The commenter stated 
that the review should not be delayed until the medical review occurs 
because such delay increases the likelihood of a claim found in error.
    Response: The DDC is responsible only for the collection of medical 
records and does not have the clinical expertise to determine the 
completeness of these records. However, the review contractor (who 
conducts the medical reviews) will notify the DDC if additional 
information is needed during the medical review, and the DDC will 
follow-up with the provider to obtain the specific information needed. 
Insufficient documentation errors are cited when the provider does not 
respond to the request for additional information or does not provide 
the additional information within 14 days of the request.
    Comment: A commenter stated that our assurances related to the 
receipt of documentation before considering an error for lack of 
documentation were insufficient. According to the commenter, it is 
unreasonable to suggest that providers will respond timely to three 
written and oral requests during a 90-day time period. The commenter 
believed the documentation/database contractor should be required to 
obtain documentation throughout the entire review year.
    Response: Our experience has shown that our provider response rate 
to requests for medical records is excellent, and that most providers 
submit records within 30 days of the original request. Therefore, we do 
not believe that the timeframe should be extended to include the entire 
review year and are not adopting this recommendation. In fact, given 
that the provider response rate is good and considering States' 
concerns with the 90-day timeframe impeding on the difference 
resolution process, we are considering reducing the timeframe, for 
example, to no later than 60 days from the date of the letter sent by 
the contractor requesting the medical records. If we decide this is 
worthwhile, we will issue a policy instruction to that effect.
    Comment: A commenter stated that, since the documentation/database 
contractor will request medical records for the PERM program from a 
provider, CMS should consider methods to minimize the duplication of 
efforts since the State will have already received documentation from 
the provider.

[[Page 50498]]

    Response: We agree that duplication of effort should be minimized 
wherever possible as long as the documentation is complete, 
comprehensive, and timely.
4. Adjustments to Claims
    Comment: A commenter requested clarification regarding whether the 
60-day adjustment timeframe pertains to managed care claims or whether 
it only applies to FFS claims.
    Response: The 60-day adjustment timeframe pertains to both FFS and 
managed care claims. Adjustments made within 60 days of the original 
paid date will be included in the review process, which will consider 
the net amount paid (original paid amount with additions and 
subtractions due to adjustments that occurred within 60 days) in 
calculating the error rate.
    States will submit adjustments for managed care payments selected 
in the random sample each quarter. These may include retroactive rate 
changes, rate cell assignment corrections, and takebacks for 
beneficiaries who lost eligibility.
    Note that, while States may have policies that allow adjustments to 
be made more than 60 days after the original paid date, only the 
adjustments made within 60 days are considered for PERM purposes.
    Comment: Several commenters expressed concern that Sec.  
431.970(a)(8) requires States to make adjustments to managed care 
capitation claims within 60 days of the adjudication date. The 
commenters maintained that States needed a longer timeframe to 
reconcile and adjust payments before the payments were classified as 
errors. One commenter observed that its SCHIP program has a 
reconciliation process in place that makes positive and negative 
adjustments to capitation payments to health plans on a retroactive 
basis; this process takes longer than 60 days. Some other commenters 
asserted that adopting a 60-day window for adjustments is contrary to 
the time periods now allowed in many States. One of the commenters 
recommended that CMS extend the adjustment timeframe to a minimum of 4 
months.
    Response: We responded to this comment in the August 28, 2006 
interim final rule (70 FR 58260). We understand the commenter's 
concern; however, States have varying timeframes in which claims are 
adjusted, and we cannot extend the timeframe in a manner that would 
accommodate all States' practices. We noted in the August 28, 2006 
interim final rule that the 60-day timeframe was agreed upon by States 
and CMS during the development of the review methodology under the PAM 
pilot projects as a reasonable timeframe. The 60-day timeframe allows 
for claims adjustments while maintaining a timeline that also allows 
for completing the reviews and computing and reporting the error rates 
in time for inclusion in the PAR.
    If we extend the timeframe to a point beyond 60 days, we cannot be 
assured that the error rate measurement process will be completed in 
time to report the error rate. Accordingly, we are not adopting this 
recommendation.
    Comment: A commenter stated that a bottom-line error rate must net 
overpayments and underpayments as already required by the HHS Office of 
Inspector General Corporate Integrity Agreements (https://oig.hhs.gov/
fraud/cia/docs/ciafaq1.html).
    Response: OMB guidance M-06-23, published on August 10, 2006, 
states that ``incorrect amounts are overpayments and underpayments 
(including inappropriate denials or payment of services).'' OMB 
guidance further directs that the estimate of improper payments is a 
gross total of both over and under payments. The OIG guidance that the 
commenter refers to is for a different purpose and does not apply to 
PERM.
    Comment: If a claim is sampled that is a reversal of a prior claim, 
a commenter asked whether States would need to provide the original 
claim, which may have been outside the timeframe.
    Response: The State will sample original claims only because no 
stand alone adjustments to claims are included in the universe. In 
other words, the State will sample original claims only and make any 
necessary adjustments within 60 days of the paid date for the claims 
after the sample is selected. These consolidated and adjusted claims 
would then be reviewed to determine if they were correctly paid.
    Comment: A commenter asked whether adjustments to claims made 
within 60 days from the adjudication dates for the original claims or 
line items should be provided for the universe, or just for the 
selected sample.
    Response: Adjustments should be provided for the selected sample 
only.
5. Medical and Data Processing Reviews
a. Methodology
    Comment: A commenter recommended separating out claims for 
residential care services within the overall estimate of the State 
payment error rate, and suggested that CMS perform a quantitative and 
qualitative analysis to determine the underlying reasons for the 
payment errors in this category through surveys. CMS could then utilize 
the data to correct the errors by giving States and providers 
additional training where needed.
    Response: Although we appreciate the commenter's recommendation, we