Statement by
Charles Johnson
Assistant Secretary for Budget, Technology and Finance
U.S. Department of Health and Human Services

HHS Improper Payment Initiatives

Subcommittee on Federal Financial Management, Government Information, and International Security
Committee on Homeland Security and Government Affairs
United States Senate

Thursday, March 9, 2006

Good afternoon Dr. Coburn and distinguished Subcommittee members. Thank you for inviting me before you today. It is a pleasure and honor for me to have the opportunity to speak on the U.S. Department of Health and Human Services (HHS or the Department) improper payment initiatives.

The Department is firmly committed to ensuring the highest measure of accountability to the American people. With the size and scope of HHS programs, we know that it is critical to prioritize, and be aggressive in our activities to identify and take action to reduce improper payments. Over the past several years, we have had many successes and accomplishments in this area. I am pleased to share some of these with you today as well as some of the challenges we face.

As required under the Improper Payments Information Act of 2002 (IPIA) and related guidance issued by the Office of Management and Budget (OMB), the Department is estimating, or in the process of developing or implementing methodologies to estimate improper payments, for seven of its programs: Medicare, Medicaid, State Children's Health Insurance Program (SCHIP), Temporary Assistance for Needy Families (TANF), Foster Care, Head Start and Child Care. These seven programs account for close to 90 percent of HHS' $640 billion total estimated FY 2006 outlays. In terms of both size and potential for growth, the risk and impact of improper payments is greatest for the two HHS programs which account for 80 percent of the total outlays - Medicare and Medicaid. My testimony here today is focused on improper payments. Those cases involving fraud are referred to the HHS Office of the Inspector General (OIG) and prosecuted by the Department of Justice which provides an important deterrent to fraudulent payment schemes.

The Department's largest program, Medicare, accounts for close to 50 percent of the Department's outlays. Medicare is a Federal health insurance program administered by HHS that provides medical insurance to 42 million people. The majority of Medicare spending is for fee-for-service (FFS) hospital and physician services. The FFS component of Medicare covers a wide range of other items and services, including home health care, ambulance services, medical equipment, and preventive services. The HHS Centers for Medicare and Medicaid Services (CMS) administers the Medicare FFS claims processing and payment systems through contracts with Carriers, Durable Medical Equipment Medicare Administrative Contractors (formerly called Regional Carriers (DMERCs)), Fiscal Intermediaries (FI), and Quality Improvement Organizations (QIOs). These entities review claims submitted by providers to ensure payments are made only for medically necessary services covered by Medicare for eligible individuals. HHS estimates that the contractors processed over one billion claims (1.156 billion claims) from providers, physicians, and suppliers for items and services that Medicare covers.

In 1996, HHS' Office of the Inspector General (OIG) began estimating improper payments in the Medicare FFS program as part of the financial statement audit required by the Chief Financial Officer's Act of 1990. The OIG produced FFS error rates from FYs 1996 - 2002. Beginning in FY 2003, CMS, working with the OIG, implemented a more robust process - the Comprehensive Error Rate Testing (CERT) program - to assess and measure improper payments in the Medicare FFS program. The CERT program not only produces a national paid claims error rate but also provides very specific improper payment rates, including contractor-specific improper payment rates which measure the accuracy of our claims processors; provider-type specific improper payment rates which measure how well the providers who care for our beneficiaries are preparing and submitting claims to the Program; and other management related information which provides insight into payment errors by region and reason. The Medicaid FFS improper payment estimate is derived from two programs; the Comprehensive Error Rate Testing (CERT) Program and Hospital Payment Monitoring Program (HPMP). Each component represents about 50 percent of the total FFS Medicare payments. The CERT Program has provided HHS with a powerful tool to identify problems in the claims process and address these problems through specific corrective action plans.

In FY 2005, HHS reported a Medicare FFS paid claims error rate of 5.2 percent; a rate that is significantly lower than the 10.1 percent rate reported in FY 2004. The significant drop in the rate is primarily attributable to the aggressive measures that were taken by the Department to ensure that necessary documentation was in place to support payment claims.

The CERT and HPMP statistical methodologies that HHS uses to calculate the Medicare national FFS error rate were reviewed by PricewaterhouseCoopers, LLP (PwC) in FY 2004. As a result of the review, PwC reported the "fee-for-service error rate to be statistically valid." In addition, reviews were done by the Department's OIG in FY 2005 and GAO is in the process of completing a review of the methodology as required under the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA).

HHS program integrity activities are primarily funded through the Medicare Integrity Program (MIP), established by the Health Insurance Portability and Accountability Act of 1996. The MIP includes medical review and benefit integrity activities, provider education and training, Medicare Secondary Payer, and provider audits. HHS overall program integrity efforts were supplemented by funding from HHS program management account and other funds made available from the Health Care Fraud and Abuse Control (HCFAC) account. Additionally, new Medicare contractor reform legislation enacted through the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA), will further enhance MIP's effectiveness.

The Administration's budget request for FY 2007 provides new resources for reducing improper payments. The budget includes $1.1 billion from the Health Care Fraud and Abuse Control account to fight improper Medicare and Medicaid payments. To supplement these efforts, the Budget requests $118 million for efforts to protect the new Medicare prescription drug benefit and the MA program against fraud, waste, and error, as well as reduce errors in Medicaid. These funds are part of a Government-wide proposal to fund program integrity activities through a discretionary cap adjustment.

While CERT and HPMP have been useful for guiding our efforts in the Medicare FFS program, they do not provide a measure for payments in Medicare Advantage or the Medicare Prescription Drug Benefit Program. These programs added by the MMA represent about 18% of Medicare benefits outlays in FY 2006, and will grow in future years. The Department is in the process of evaluating how to best address improper payments in these programs.

In 2001, using HCFAC funding, HHS embarked on its first Medi-Medi project with the State of California. This project allowed HHS to match Medicare and Medicaid data to detect fraudulent patterns that may not be evident when billings for either program are viewed in isolation. Since that time, nine States have participated. Since inception, the Medi-Medi projects have yielded 335 investigations with an estimated $182 million dollars at risk.

The Department's second largest program, Medicaid, accounts for over 30 percent of Department outlays. Unlike Medicare, it is administered primarily by State Governments. While the Federal Government provides financial matching payments to the States, each State is responsible for overseeing its Medicaid Program, and each State essentially designs and runs its own program within the Federal structure. The Federal Government pays the States a portion of their costs through a statutorily determined matching rate called the Federal Medical Assistance Percentage, or FMAP, that currently ranges between 50 and 77 percent. In FY 2006, total Medicaid expenditures - those that include both Federal and State contributions - are estimated to be approximately $340 billion.

In FY 2000, HHS adopted a Government Performance and Results Act (GPRA) goal to explore the feasibility of developing a methodology to estimate improper payments in the Medicaid Program. Beginning in 2001, HHS formally solicited States to participate in the development of a model to estimate payment accuracy. Only three States, Illinois, Texas, and Kansas, had attempted to estimate payment error in their respective State Medicaid Programs prior to HHS initiating pilot projects.

From FYs 2002-2005, HHS conducted the Payment Accuracy Measurement (PAM) and Payment Error Rate Measurement (PERM) pilot projects with extensive collaboration from participating States to determine a systematic means of measuring payment errors at the State and national levels. From these pilot projects, HHS was able to develop a methodology to estimate a State-specific payment error rate that would be the basis for the national Medicaid error rate as well as the State Children's Health Insurance Program (SCHIP).

In FY 2006, contractors will measure a national Medicaid FFS error rate in 17 States based on medical reviews and data processing reviews. In FYs 2007 and 2008, contractors will measure national Medicaid and SCHIP FFS and managed care (MC) payments in 17 randomly selected States, and the States will measure eligibility payment errors. Comprehensive Medicaid and SCHIP error rates (MC, FFS and eligibility) will be reported in the FY 2008 PAR.

The TANF program provides a capped pre-appropriated annual block grant of approximately $16.7 billion to States, Territories and eligible Tribal programs to help families transition from welfare to self-sufficiency. In the past several years, HHS has worked toward identifying strategies for estimating payment errors in the TANF Program. Four different activities were identified to assist in efforts to reduce the occurrence of improper payments in the TANF Program. These activities and related actions taken include:

  • HHS is soliciting information from States on their practices for identifying and reducing improper payments in the TANF Program. HHS developed a survey instrument to solicit information on State systems and practices for identifying and reducing improper payments in the TANF Program that will be placed on a website for information sharing among the States;
  • HHS is conducting an improper payments demonstration project with volunteer States in which the States undergo a more in-depth review of TANF expenditures during the OMB Circular A-133, Audits of States, Local Governments, and Non-Profit Organizations, audit process. The objective of the pilot is to explore the viability of estimating improper payments using the A-133 audit process. HHS obtained agreement from one State (Alabama) to participate in the A-133 audit pilot project. In the expanded audit, the auditors used a statistical sample of a fixed size for a test of controls (attribute sampling method). The auditors reviewed 208 TANF cases to achieve a 95 percent confidence level with an expected deviation rate of 2.25 percent. The auditors reported an overall case error rate of 20 percent and a payment error rate of 3.9 percent. HHS contacted six States to increase the number of States participating in the A-133 pilot in FY 2006. Of the six States contacted, only three States agreed to participate. HHS will report on the results of the audits in these three states in the FY 2006 PAR;
  • HHS initiated various activities to improve data match capability and increase State utilization of the Public Assistance Reporting Information System (PARIS). PARIS provides a Federal computer matching capability to assist State Public Assistance Agency (SPAA) efforts to validate client-reported information and identify potential improper payments (using client social security numbers) in the Medicaid, TANF and Food Stamps Programs. PARIS includes the Veterans Administration (VA) match and a VA spousal match; a Department of
  • Defense/Office of Personnel Management match (active and retired military personnel and Federal employees); and an interstate match (duplicate payments made to the same client in more than one State). Every quarter, PARIS member States voluntarily choose whether, and in which match to participate (at no charge to them). The more States that join and conduct matches under PARIS, the wider the net of potential matches of information becomes available to PARIS member States to validate public assistance program client-reported information and identify potential improper payments - especially under the interstate match. HHS also engaged in a number of activities to improve data match capability and usefulness and increase State utilization of PARIS and will be continuing to work on expanding State participation and improving PARIS capability in FY 2006; and
  • HHS is continuing to expand State access to the National Directory of New Hires (NDNH). The NDNH offers solutions to the prevalent under-detection by States and reporting of employment of TANF recipients. The NDNH was authorized under the welfare reform legislation to provide a national database of employment information for the purpose of collecting child support payments. The NDNH contains three database components: 1) new hires - information on new employees (filled out W-4 data); 2) quarterly wage data which includes information on individual employees from the records of State workforce and Federal agencies; and 3) unemployment compensation. HHS has initiated a demonstration project to provide State TANF agencies direct access to match their TANF caseloads against the databases. This effort began with a pilot effort in the District of Columbia (DC). In the DC pilot, 33 percent of the individuals submitted were identified as employed by the match, and over 81 percent of those identified were verified as actually being employed. The vast majority of these recipients were not known to be employed by the State TANF agency. In FY 2005, all State TANF agencies were given access to the NDNH. To encourage use of the NDNH to carry out program responsibilities, HHS has provided States access to conduct up to 12 matches (one per month) against the New Hires (W-4 data) database in FY 2006. Since July 2005, 30 States, DC and Puerto Rico have conducted matches. Together, these States and Territories account for 82 percent of the TANF caseload. During FY 2006, HHS will continue working with the States.

Although HHS is engaging in many activities which have been quite successful in identifying improper payments, HHS has not yet identified an efficient and effective approach for determining an estimate of improper payments in the TANF Program. One of our most significant challenges has been the flexibility that States have in the design and administration of the program. Also, there are statutory limitations with regard to the information that HHS can request of States. HHS is in the process of considering the work that has been done thus far and will continue to work toward formulating a feasible and detailed plan for estimating payment errors in TANF.

The Head Start Program provides grants to local public and non-profit agencies to provide comprehensive child development services to children and families, primarily preschoolers from low-income families. Head Start regulations allow Head Start programs to serve up to 10 percent of their enrolled children (49 percent in certain situations for tribal Head Start programs) from families who do not meet Head Start income requirements. Under Head Start legislation, grantees are required to be monitored at least once every three years. In FY 2004, HHS developed a methodology for estimating a national Head Start payment error rate building on the required review process. HHS has reported Head Start payment error rates in FY 2004 (3.9 percent) and FY 2005 (1.6 percent).

The Foster Care Program is designed to help States provide safe, appropriate, 24-hour, substitute care for children who are under the jurisdiction of the administering State agency and who need temporary placement and care outside their homes. Under the regulatory review promulgated at 45 CFR 1356.71, primary reviews are conducted in each State every three years by teams who review 80 cases selected from the State's title IV-E foster care population. These reviews are intended to recover title IV-E funds claimed by States for ineligible cases and, in conjunction with the required program improvement plan (PIP) for those States determined to be non-compliant, to help change their behavior so that subsequent reviews will result in lower error rates.

HHS developed a methodology for estimating a national payment error rate for the title IV-E Foster Care Program using data gathered in the eligibility reviews conducted in FY 2001 - 2004. The FY 2004 error rate was 10.33 percent and the FY 2005 final error rate was 8.6 percent.

HHS has begun measuring underpayments in the reviews that are being conducted in FY 2006. In the coming year, HHS will continue to measure error cases and begin implementing its plan to measure Foster Care administrative cost payment errors.

The Child Care and Development Fund (CCDF) is a block grant composed of three distinct funding elements (mandatory, discretionary and matching) authorized in two different statutes. During FY 2003, HHS began to work toward identifying strategies for estimating payment errors in CCDF . In FY 2004, HHS initiated an improper payment pilot project to assess the efforts of eleven States to prevent and reduce improper payments in their child care programs and to explore feasible strategies to measure and estimate improper payments for the program. HHS expanded State participation in the pilot project from eleven to eighteen States in FY 2005 and continued to work on a strategy for determining a payment error rate in the CCDF. Further, HHS partnered with Regional and State staff to test an error rate methodology in four States focused on the client eligibility process.

HHS drafted a report of the findings which includes a preliminary error rate calculated for each of the four States and an estimated analysis of the cost incurred by each State. HHS also developed a survey instrument to solicit information on a voluntary basis from States on State systems and practices for identifying and reducing improper payments in the CCDF.

CCDF gives the States flexibility in the design and administration of the Program which has presented challenges in developing a model or methodology that can be used by all States. HHS is developing a plan for applying the error rate methodology that was tested in the four States in FY 2005, to all the States over time. This methodology focuses on client eligibility and involves an intensive case review process to identify cases with errors, cases with improper payments, percentages of payments made in error, average amounts of improper payments, and minimum and maximum amounts of improper payments. The field work also raised questions about the need for regulatory or other policy changes to support the State-by-State error rate strategy. These policy questions are currently being considered within the Department.

Dr. Coburn, and Subcommittee member, in conclusion, HHS has had numerous accomplishments and successes in its improper payment activities. In our largest program, Medicare, we are estimating improper payments and seeing the results of the corrective actions we have taken in seeing the rate drop significantly. In Medicare, Head Start and Foster Care Programs we experienced a decrease in improper payments through identification and implementation of appropriate corrective action. In two other programs, Head Start and Foster Care, we have achieved efficiencies in utilizing reviews required by legislation or regulation in developing our methodologies for determining estimates in these programs. In our second largest, Medicaid, we have developed and are working on implementing a plan for estimating improper payments. In the two programs where we have not been able to develop methodologies for estimating improper payments, TANF and Child Care, we are engaging in activities that comply with the intent of the IPIA in identifying and reducing improper payments. Our data matches and pilot activities have not only been successful in identifying and reducing improper payments, they have allowed us to build strong partnerships with the States in our endeavors to reduce improper payments. In the coming months, we will continue to work toward achieving compliance with the IPIA in overcoming the challenges we face in our TANF and CCDF and in implementing our plans for estimating improper payments in the Medicaid and SCHIP Programs.

Thank you again for this opportunity to talk about the Department's improper payment initiatives. At this time, I will be pleased to answer any questions.

Last Revised: March 10, 2006