Office of Investigations and Enforcement (OIE)
Medicaid Fraud and Abuse
SB 1587 Biennial Report
The Effectiveness of Alternative methods for Proof of Eligibility in the Medicaid Program
November 2001
(printable document - adobe acrobat/pdf format)
download required viewer

Introduction

Section 7(d) of Senate Bill 1587, as enacted by the 76th Legislature, requires that the Health and Human Services Commission (HHSC) provide a biennial report to the Legislature regarding "the effectiveness of any alternative method for proof of eligibility under the state Medicaid program implemented by the Texas Department of Human Services in reducing incidences of fraudulent claims of eligibility." This document provides the contribution of the Texas Department of Human Services (TDHS) toward the Section 7(d) report.

For the past several years, Texas and the nation have been engaged in an extensive discussion about Medicaid fraud. However, the terms of the debate have focused almost entirely on Medicaid provider fraud. Comparatively little exploration of Medicaid eligibility fraud has occurred. Where client Medicaid fraud has been discussed, it is usually in the context of schemes in which a provider and one or more recipients collude to obtain Medicaid reimbursement. As the Robert Wood Johnson Foundation noted in the June 1999 issue of its newsletter, State Initiatives, "Most fraud cases that are pursued involve fraudulent providers, because they offer the greatest potential for financial recovery." The Foundation also noted that some jurisdictions experience difficulty obtaining indictments, convictions, and repayment for recipient Medicaid fraud.

Consequently, there is no clear picture regarding how Medicaid client fraud occurs or how frequently it takes place. This document seeks to provide a comprehensive look at the issues, in the context of the current system and the enhancements recommended by the Texas Comptroller of Public Accounts in Texas Electronic Services Delivery: Final Report. Those issues include the state of Medicaid client investigation in Texas and how the existing technologies for replacing it could potentially affect the detection, investigation and incidence of Medicaid client fraud.

top of page

The current system

Under the current system of eligibility notification, TDHS mails a notice, known as Medicaid Identification (Form 3087), to the address of certified cases. For example, the July monthly notice was mailed on June 23, 2001 to 1,716,054 clients, representing 1,284,643 households. There are also supplementary mailings on a weekly basis for the purpose of providing proof of coverage to cases certified after the monthly mailing. In May and June mailings, an additional 153,752 clients, representing 112,732 households received an initial proof of eligibility in supplemental mailings.

Form 3087 is a paper form that a client can use to prove eligibility. It contains information regarding the number of individuals on each case up to nine. Up to three Form 3087s can be mailed in a single envelope. The Form 3087 contains a significant amount of information designed to inform clients and providers of the type of services for which the individual has Medicaid coverage.

All recipients of Medicaid services, including long-term care, receive Form 3087, with the exception of several small programs, which represent partial benefits. Those partial Medicaid programs are the following: 1929b waiver programs (in home services); Specified Low-Income Medicare Beneficiary (premium payments only); and residents of State Schools.

Specifically, Form 3087 contains the following types of information:

• The recipient's identifying information, including name, date of birth, sex, the Medicaid identification number, the case category, and case number;
• The dates on which the recipient became eligible, on which the form becomes invalid, and on which it was printed;
• Information pertaining to the type of program in which the recipient is enrolled (TP), and whether eligibility is determined under TDHS or SSI procedures;
• The services for which the recipient is eligible, such as hospice, or a limitation to emergency care only, and whether the recipient is presumptively eligible, limiting services to medically necessary outpatient services;
• Whether or not the recipient is limited to services from a single physician or pharmacy;
• Whether TDHS is providing coverage of the Medicaid deductible and coinsurance,
• Whether the recipient is enrolled in a Medicaid managed care program, and, if so, the name of the provider;
• The bank identification number for the vendor drug program;
• Whether the recipient has private or Medicare insurance, including the Medicare number, if known; and
• Information about Texas Healthy Steps' Medical and Dental check-ups for recipients younger than 21 years of age.

Form 3087 also proves eligibility for the Vendor Drug program. In that program, an authorized representative may use the Form 3087 to obtain covered prescriptions on the client's behalf.

top of page

SB 1587 and Modifications to Form 3087

Section 7(a) of Senate Bill 1587 required TDHS to develop a Medicaid eligibility letter that is not easily duplicated before October 1, 2000. In response to this requirement, TDHS determined that the most cost-effective manner of complying with the legislation would be to print Form 3087 on a type of paper with a watermark that could not be reproduced. TDHS obtained samples of paper with watermarks for review. The agency selected a light watermark in the form of a state seal on a kind of paper that is considered "safety paper," meaning that attempts to alter the print with eradicating ink or bleach will not work. The first mail-out of this new paper took place around April 1, 2001. Providers were informed of the change in the proof of eligibility through mailings from the National Heritage Insurance Corporation.

top of page

The Complexity of the Medicaid Programs

TDHS already operates an electronic benefits transfer (EBT) system for Food Stamps and Temporary Assistance for Needy Families (TANF). The technology used is a magnetic-stripe card system similar to that used by credit card companies. The only information stored on the card is a unique alpha-numeric sequence that allows access to a recipient's benefit amounts, information which is stored in the vendor's host automation system. The unique alphanumeric system, in combination with a PIN number, allows access to the benefit amount information stored by the vendor's automation. Transactions within the recipient's account, such as deposits, debits, and, in the case of returns of merchandise, credits, are handled within the vendor's host system as it communicates with the retailer's point-of-sale device.

The current EBT system handles two programs: Food Stamps and TANF. Both programs are, by their nature, economic assistance programs in which the key information is the amount of benefits, as modified by transactions. The key difference is the purpose for which the benefits can be used.

Medicaid is substantially different because it provides insurance coverage for services rendered, rather than a cash amount. The Medicaid program actually includes 24 separate programs. Eligibility criteria and the associated medical services vary substantially between those programs. The federal framework governing Medicaid provides states with authority to choose certain options and request waivers to add programs or operate them in additional ways. The result is a degree of complexity that can be difficult to understand for recipients and providers. That complexity and difficulty must be taken into account in the development of any system for providing recipients with proof of eligibility.

However, because of the way the EBT card was developed, in which the card itself does not carry benefit amount information, it is conceivable that Medicaid could be added to the same card. That, in fact, is the recommendation of the Texas Electronic Services Delivery: Final Report. The host automation system would be responsible for maintaining information regarding recipient identification, program type, eligibility, range of services and even services rendered.

top of page

Medicaid Recipient Fraud

Medicaid Eligibility Fraud

The language in Section 7(d) clearly focuses on fraud committed for the purpose of obtaining Medicaid eligibility, but that is only a portion of the recipient fraud problem. States generally think of eligibility fraud in the context of the Food Stamp and TANF programs, in which a recipient provides false information or withholds information, for the purpose of obtaining benefits to which the household is not entitled. TANF funds are not restricted by use, so fraud in that program is limited to misrepresentations for the purpose of obtaining eligibility and do not extend to the misuse of benefits. Food Stamps transactions are limited to purchases of food, so misuse can occur if a recipient uses those benefits to obtain something that cannot be bought under program rules. Generally, inappropriate use of Food Stamp benefits takes the form of trafficking: the sale of benefits by a recipient, who is often eligible for the program, to a retailer, in exchange for cash.

Discussions of recipient Medicaid fraud are usually conducted under assumptions borrowed from the somewhat different universe of cash benefit eligibility fraud. In fact, the overwhelming proportion of eligibility fraud identified and referred to prosecutors by TDHS' Office of Inspector General (OIG) is associated with cases in which a recipient has obtained TANF benefits by fraud and is therefore eligible for Medicaid. In those cases, the state has paid Medicaid premiums and, in some cases, reimbursed providers for services rendered. Depending on the circumstances, either or both of those amounts may be considered fraudulent. In general, if OIG investigators determine that fraud has occurred in the Food Stamp, TANF or Medicaid, they check all three programs.

For example, in FY 1999, all but 55 of the 957 Medicaid cases (84 percent) investigated by OIG were for public assistance-related fraud. In FY 2000, all but 137 of the 1,438 investigations of fraud in Medicaid programs, or 90 percent, were associated with public assistance-related fraud.

The fact that so much of OIG's current investigative workload is associated with public assistance-related Medicaid does not mean that fraud does not exist in the Medicaid-only programs, either acute care or long-term care. Since the Medicaid program provides a wider variety of benefits, such as acute care services, prescription drugs, and long-term care services, both the motivation and the methods for committing fraud are different. Consequently, to be effective, the analysis, means of detection, and investigative procedures cannot be based solely on assumptions related to cash benefit eligibility fraud.

Types of Fraud Cases

In its analysis of Medicaid recipient fraud, OIG has identified the following scenarios for fraud, including both eligibility and post-eligibility, based on experience:

Acute Care Eligibility Fraud: OIG has discovered cases in which recipients gave false information to Texas Works Advisors for the purpose of obtaining acute care coverage. OIG has knowledge of cases in which individuals have falsified information to caseworkers in order to qualify for Medicaid because they are uninsured or because they are paying premiums for dependents. The advent of the Children's Health Insurance Program (CHIP) may reduce some of the motivation for this kind of fraud. Based on OIG's records, this does not appear to be as widespread as eligibility fraud in the public assistance programs.

Long-term Care Eligibility Fraud: OIG has also investigated cases in which individuals hide family assets in order for a member to qualify for nursing home care under the Medicaid program. Investigations of this kind of fraud are infrequent, in part because nursing facility caseloads are much lower than the public assistance caseloads. It may also occur with less frequency than investigations of public assistance fraud because the number of referrals received from the public is lower.

Duplication of Proof of Eligibility: Section 7 of SB 1587 is based on the perception that duplication of Form 3087 is a serious problem. OIG has not identified this as a serious problem in Texas, although in at least one state (Alabama) an individual created fraudulent proof of eligibility and gave them to a number of recipients.

Medicaid Card-Sharing: This type of fraud takes place when a legitimately eligible recipient shares a card with another individual. There are cases in which the recipient is trying to assist a friend; there have also been instances in which the proof of eligibility has been sold to another individual for purposes of obtaining acute care services.

Vendor Drugs: This type of fraud takes place when an individual uses the Medicaid proof of eligibility to obtain drugs, which are later sold to other individuals. There are variants of this scenario which involve the collusion of either the prescribing physician or the pharmacist.

Collusion: There are cases in which providers may pay a recipient in order to bill for services that were not performed. This may be especially true for providers working with merchandise, including pharmaceutical drugs or durable medical equipment, such as wheelchairs.

Sources of Information about Potential Fraud Cases

This issue cannot be discussed without an understanding of where OIG obtains information regarding fraud. For cases of Food Stamp and TANF fraud (which can include Medicaid eligibility fraud), the primary sources are the public at large and Texas Works Advisors. In recent years, TDHS has taken advantage of new technology, under the authority of state or federal legislation, to locate individuals receiving benefits who are in fact ineligible, such as inmates of the institutions of the Texas Department of Criminal Justice or those who have made duplicate applications, either within Texas or in this and neighboring states.

For Medicaid acute care-only cases, the majority of referrals do not come from either Texas Works Advisors or the public. Instead, the majority of tips received by OIG come from providers - principally physicians, pharmacists and hospitals. OIG receives few referrals regarding long term care Medicaid fraud.

top of page

Baselines

One of the problems with measuring Medicaid fraud incidence or performance is the lack of any baseline from which to analyze change. This portion of the document discusses several measures.

OIG Medicaid Investigations, FY 1997-FY 2000

OIG investigations of Medicaid recipients have increased dramatically over the past few years. This is probably due to a number of reasons, including increased focus on Medicaid fraud in general, better access to technology that allows the agency to identify and prove Medicaid recipient fraud, and growth in both the Medicaid caseload and the ranks of the uninsured. It does not necessarily reflect a growth in the incidence of Medicaid recipient fraud.

Growth in Medicaid Recipient Fraud Investigations, FY 1997-FY 2000

The state only collects Medicaid in cases in which fraud has been established or in which a client voluntarily pays restitution.

Quality Control Studies

Historically, states have found that Medicaid public assistance cases have low error rates, probably because of higher eligibility income limits and the general lack of resources in these households. Because of that, Texas, with the approval of the Center for Medicare and Medicaid Services (CMS), formerly known as the Health Care Financing Administration, has not conducted overall error rate reviews of the entire Medicaid program since FFY 1994. Instead, with the encouragement of CMS, Texas has developed a series of plans to review specific Medicaid programs. The state Medicaid quality control plans must be approved by CMS.

Since FFY 1997, Medicaid quality control activities have focused on several types of Long-Term Care Medicaid cases. For the past two years, TDHS has been reviewing Medicare Savings Program cases, Qualified Medicare Beneficiary (QMB), Specified Low-Income Medicare Beneficiaries (SLMB), and Qualifying Individuals (QI 1). These are community-based programs for the Aged and Disabled population. Eligibility for these cases is determined by regional TDHS staff using a streamlined application and verification procedure that includes a telephone interview rather than a face-to-face interview in the majority of the cases. The most recent studies, for QMB, SLMB, and QI-I, showed that in FY 1999, 3.4 percent of the quality control sample was coded "fraud" and a referral was made to OIG (39 cases out of 1198). In FFY 2000, 2.1 percent of the quality control sample was coded "fraud" and a referral was made to OIG (20 cases out of 954).

These figures pertain to a specific long-term care population, and cannot be used to make generalizations about fraud activity in the Medicaid-public assistance population.

Medicaid Fraud Study

In 1998, the Office of the Comptroller conducted the first Health Care Fraud Study, under the provisions of Section 403.028 of the Texas Government Code. That study focused on eligibility fraud. The Office of the Comptroller contracted with a vendor that attempted to contact a sample of Medicaid recipients by telephone. OIG visited the residences on record of those Medicaid recipients who the vendor could not contact. This study, after consultation with the State Auditor's Office, HHSC and TDHS, reported that six percent of the sample could not be located.

HHSC/TDHS Utilization Study

In FY 2000, HHSC and OIG cooperated on a study designed to determine the incidence of fraud in cases where there was a child in the family but no Medicaid utilization was reported. The theory behind the study was that children generally require at least one physician visit per year. HHSC provided utilization data, and, by cross-referencing it with TDHS case files, approximately 800 cases were identified where no Medicaid utilization had taken place in FY 1999 despite the presence of a child in the family. OIG fully investigated 32 cases, chosen on a random basis, but did not find any fraud in the sample.

A follow-up effort focused on an additional thirty cases in which an adult member of a Medicaid household received health care services, but no children did. OIG investigated each of these cases to determine whether the children existed. OIG found that all the children existed and resided with their families, and therefore that no fraud had occurred.

Amendment to the Medicaid Fraud and Abuse Detection System Contract

Section 4 of SB 1587 provided HHSC with the authority to "contract with a contractor who specializes in developing technology capable of identifying patterns of fraud exhibited by Medicaid recipients to: (1) develop and implement the fraud detection technology; and (2) determine if a pattern of fraud by Medicaid recipients is present in the recipients' eligibility files maintained by the Texas Department of Human Services." Section 5 of the same bill required HHSC, in cooperation with OIG,. to "study and consider for implementation fraud detection technology or any other technology that can identify information in the eligibility file of a Medicaid recipient that indicates potential fraud and the need for further investigation."

Pursuant to that legislation, OIG worked with HHSC to evaluate alternatives. OIG determined that a thorough evaluation of Medicaid recipient fraud, one which takes into account the types of fraud that occur, the reasons why it occurs, and the high proportion of referrals that come from providers, must conclude that eligibility fraud is only one dimension of the problem. In order to obtain that multi-dimensional view of the Medicaid fraud problem, OIG recommended that the most cost-effective and efficient method of developing an automated system to examine eligibility fraud would be to amend the Medicaid Fraud and Detection System contract between HHSC, EDS and HNC to include a Medicaid recipient fraud component.

That contract amendment was signed on February 1, 2001. The project will last for six months, with an option to renew for an additional year, provided that the contract between HHSC, EDS and HNC is renewed. HNC is working with OIG to identify suspicious patterns in eligibility and utilization data that may indicate that recipient fraud is occurring. HNC has already provided one report, which OIG is currently evaluating. Further analysis will include a review of the claims data and continued review of the report data. OIG has identified funds within its budget to extend the contract amendment for the option year. Further work will depend on the structure of any future Medicaid fraud systems contracts.

top of page

Alternative Means of Proving Eligibility

The Comptroller's report, Texas Electronic Services Delivery: Final Report, identified the following alternative types of proof of eligibility.

Magnetic-Stripe Cards

Of all the potential replacements for the paper proof of eligibility, the most familiar to the general public is the magnetic-stripe card, which is used for almost all credit and debit transactions and is used by the Lone Star Card. Such technology is capable of retaining only non-changing information, such as name and account number, and not information that changes, such as eligibility status. In other words, it requires that the card communicate on-line with a host automation system that can manage the amount of detail currently spelled out in the Form 3087. The warehousing of information in host automation would be useful in detecting and investigating fraud, provided that it includes utilization data and that OIG has access to it.

Texas Electronic Services Delivery: Final Report notes that Magnetic Stripe Cards are easier to manufacture fraudulently than other forms of proving eligibility. However, the Lone Star Card uses the same technology and OIG is not aware of any cases of fraud arising from the use of a counterfeit card.

Smart Cards

Smart cards include an embedded computer chip. Some are limited to memory; others include a micro-processor. Some cards can download data and even use public key encryption and digital signatures. Others have radio transmitters that work without requiring a card reader. Smart cards are used primarily in Europe; currently, there is limited infrastructure in the United States to support the technology.

If information is stored on the card, rather than on host automation, OIG will have less chance of access to the eligibility and utilization data that helps with the identification and investigation of fraud. However, smart cards provide a higher degree of security; they are harder to counterfeit and they can carry a large amount of identification and eligibility information. Nevertheless, there are significant obstacles to their implementation at present, including a significantly higher cost and the lack of infrastructure and lack of public awareness built on decades of credit card usage.

Bar Coded Cards

This option is similar to the bar code system. The public is very familiar with its use, but the Comptroller reports that the level of security against fraud is very low. In addition, the amount of information that they can store is very small.

Wireless Card-Reader Technology

Wireless-card technology represents an advance on EBT technology that addresses the problem of needing an on-line connection by using cellular or other wireless connections. In terms of fraud prevention, it probably does not represent an advance over the magnetic stripe card. It may provide additional convenience for clients in cases where the online system is temporarily unavailable, but Texas Electronic Services Delivery: Final Report recommends addressing this problem through an Automated Voice Recognition (AVR) system.

Biometrics

Biometric systems link personal biological identifiers, such as finger-prints, retinal scans, voice verification, signature verification, hand geometry, or facial recognition. Biometic identifiers provide a very high level of security and deterrence but are expensive.

Legislation from the 77th Legislature

Medicaid eligibility was a key issue during the 77th Legislature. The legislative process, which included consideration of reports from the Blue Ribbon Task Force on the Uninsured and the Office of the Comptroller and of extensive testimony from TDHS staff, resulted in the adoption of Senate Bill 43. This legislation is intended to address concerns about the complexity of the Medicaid application process. SB 43 provides for the following changes to the Medicaid eligibility process:

• The application form and procedures for children's Medicaid are required to be the same as for the Children's Health Insurance Program (CHIP).
• The face-to-face interview requirement is eliminated, and mail applications and recertification by mail or phone are allowed.
• Other health and human services agencies, such as hospitals, are allowed to accept applications for children's health care.
• Documentation and verification procedures (including assets) are required to be the same as for CHIP.
• TDHS is required to develop procedures to ensure that information regarding a child who will be denied Medicaid because of income, assets or resources, but will be eligible for CHIP, is transmitted to CHIP promptly.
• Children who lose eligibility for Medicaid can make the transition to CHIP without interruption in coverage.
• TDHS is required to develop procedures for contacting and informing a parent/caretaker of a child who will be denied Medicaid because of failure to keep an appointment or failure to provide information or other procedural reason regarding the need to recertify and the availability of medical coverage.
• A health care orientation is required for a parent or guardian, to be provided either by TDHS or a health care provider; compliance with this provision qualifies the child for continuous eligibility, and non-compliance results in a requirement for a face-to-face recertification.
• Compliance with Texas Health Steps is required; compliance with this provision qualifies the child for continuous eligibility, and non-compliance results in a requirement for a face-to-face recertification.

Impact of SB 43

TDHS does not anticipate any appreciable impact from the enactment of SB 43 on either the choice and implementation schedule of any alternative means of proving eligibility for Medicaid. The main provisions of the bill affect households in which there are no adult Medicaid clients. Households without adult members form a relatively small proportion of the overall caseload and, in the assessment of OIG, are less likely to be fraudulent.

top of page

Conclusion

SB 1587 requires that this report consider the effectiveness of alternative methods of proof of Medicaid eligibility in reducing the incidence of fraudulent claims. A bar code provides the lowest level of security and is not seriously under consideration for implementation. The greatest level of security is provided by biometric identifiers and smart cards. Wireless card-readers and magnetic-stripe cards provide an intermediate level of security, which probably meets the threshold needed for the Medicaid-ID.

Each of the methods reviewed in Texas Electronic Services Delivery: Final Report represents a significant improvement over the current method of proving eligibility in terms of security, especially when that concern is articulated in terms of the ease of duplication. For that reason, other important considerations, such as cost-effectiveness, acceptance by the provider and client community, and compatibility with a back-up system to be used at times when the electronic system is unavailable should be considered.

Texas Electronic Services Delivery: Final Report details a plan for the development of a Medicaid-ID by adapting the current EBT card, which uses magnetic-strip technology. Heads of household would receive multi-program cards; family members would receive cards that would provide them with proof of Medicaid eligibility. This proposal satisfies the requirements for an electronic means of verification. A magnetic stripe card offers a level of security sufficient for this program, and is a significant improvement over the current system. Medical providers and clients are familiar with such systems, and the report envisions a back-up AVR system.

The major concern that has arisen with regard to the report is the timing of its implementation. The targets of the project are June 2001 for planning and design; September 2002, for the pilot, with a statewide rollout projected for February 2003. This project entails both a substantial modification to the current Lone Star card and a stand-alone Medicaid-ID. TDHS agrees that the solution proposed by Texas Electronic Services Delivery: Final Report is the best and most effective solution to the problem of proof of Medicaid eligibility, but finds that the timeline is aggressive and does not allow sufficient time to achieve all the goals of the report, which include not only the Medicaid-ID, but the electronic delivery of WIC benefits and child support payments.

back to OIE home | top of page