United States General Accounting Office GAO Report to Congressional Requesters June 1997 SUPPLEMENTAL SECURITY INCOME Timely Data Could Prevent Millions in Overpayments to Nursing Home Residents GAO/HEHS-97-62 United States GAO General Accounting Office Washington, D.C. 20548 Health, Education, and Human Services Division B-271862 June 3, 1997 The Honorable Nancy L. Johnson Chairman, Subcommittee on Oversight Committee on Ways and Means House of Representatives The Honorable E. Clay Shaw, Jr. Chairman, Subcommittee on Human Resources Committee on Ways and Means House of Representatives Recent growth in the Supplemental Security Income (SSI) program and approximately $1 billion in annual overpayments to SSI recipients have increased congressional interest in ensuring that SSI recipients receive only those benefits to which they are entitled. In 1996, about 6.6 million SSI recipients received about $24 billion in federal payments and $3 billion in state supplemental payments, and the maximum monthly SSI federal benefit for eligible individuals was $470. There were about 1.8 million individuals in nursing homes and other similar institutions having their care paid by Medicaid at a cost of about $39 billion in 1995, and contrary to law, some of them were continuing to receive their full SSI benefits. This usually occurs because the Social Security Administration (SSA), which administers the SSI program, is unaware that the individuals are in Medicaid facilities. Because of your concern that some SSI recipients in nursing homes and other medical treatment institutions may receive overpayments, you asked us to determine (1) the extent of such overpayments, (2) the success or failure of SSA actions in preventing and detecting these overpayments, and (3) the methods by which SSA can better prevent such overpayments. To answer these questions, we interviewed officials from the Health Care Financing Administration (HCFA), which is the U.S. Department of Health and Human Services (HHS) agency responsible for the Medicaid program; SSA headquarters in Baltimore, Maryland; 4 SSA regional offices; and 13 field offices. We also visited with Medicaid agencies in five states (California, Florida, New York, Tennessee, and Texas) to collect information on their Medicaid data systems. In addition, we analyzed SSA data on detected overpayments caused when SSA was not notified in a timely manner of SSI recipients’ admissions to medical institutions. We also obtained Medicaid nursing home admissions data for December 1996 from Page 1 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 New York and Texas to determine the number of SSI recipients recently admitted to nursing homes in these states and potentially receiving overpayments. To demonstrate the potential of an automated information interface between SSA and state Medicaid agencies to help minimize overpayments, we obtained Tennessee state Medicaid data on admissions to nursing homes and other Medicaid facilities and matched that information against the Supplemental Security Record (SSR), SSA’s payment record for the SSI program. This showed those recipients who were residing in nursing homes but who were still receiving full SSI benefits. (See app. I for more information on our scope and methodology.) SSA estimates that overpayments to individuals in nursing homes may Results in Brief exceed $100 million annually; however, the exact extent is unknown. Despite SSA procedures to prevent overpayments, and recent legislation designed to further help prevent these overpayments to SSI recipients in nursing homes, we determined, based on SSA data, that it had detected overpayments totaling $24 million to about 31,000 recipients in fiscal year 1995. Furthermore, in two states we visited, New York and Texas, we determined that SSA may not have been aware of an additional 1,699 SSI recipients recently admitted to nursing homes during a 1-month period and potentially overpaid these individuals $515,714 in benefits during the subsequent month alone. SSA efforts to prevent these overpayments or detect them in a timely manner have had little success. In many cases, recipients or their representative payees1 did not report the change in living arrangements in a timely manner. In addition, because of other work priorities, SSA field representatives have not routinely contacted the over 23,000 U.S. nursing homes to solicit their cooperation in notifying SSA of admissions of SSI recipients, as SSA policies require. Furthermore, our analysis of SSA data shows that overpayments to SSI recipients residing in nursing homes have increased by nearly 13 percent since the October 1995 effective date of the legislation that was designed to reduce overpayments. While the effect of this legislation is difficult to determine because SSA does not have uniform systems for either following up on admission notifications from nursing homes or monitoring compliance with the law, it has likely been diminished by limited SSA actions to enforce the reporting requirement and the lack of a statutory penalty for nonreporting by nursing homes. Moreover, SSA’s other processes to detect overpayments in a timely 1 Representative payees are individuals or organizations that receive checks on behalf of SSI recipients who are unable to manage their own affairs. A representative payee is responsible for dispensing the SSI payment in a manner that is in the best interest of the recipient. Page 2 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 manner have not been effective. SSA’s redetermination process does not always provide timely identification of individuals in nursing homes because redeterminations are typically conducted only once every 6 years. Also, SSA’s annual computer match with HCFA does not contain Medicaid data from all states; does not result in timely identification of admissions; and places a substantial, manual workload on field office personnel. SSA could more quickly detect overpayments by electronically obtaining nursing home admissions data directly from states to help identify recent changes in recipients’ living arrangements. SSA could then use an automated interface to automatically adjust the benefits of SSI recipients admitted to nursing homes who are ineligible for continuation of benefits due to temporary institutionalization and prevent the occurrence of overpayments for the ensuing months. In the states we visited, we found that the state Medicaid agencies can make these data available to SSA. SSA could use an existing data exchange system with states to obtain the needed data electronically. Additionally, we found that in the interim period, while some states prepare their automated systems to make the electronic exchange of data with SSA, states could provide SSA with tapes or paper listings of this information for SSA’s use in detecting overpayments. To identify the program improvements such an automated interface could potentially produce, we conducted a joint effort with SSA and the Tennessee Department of Health Services. Through this match, SSA identified $31,000 in overpayments to individuals who had been approved for nursing home admission during February 1996 that SSA had not previously detected. Furthermore, by identifying SSI recipients in nursing homes sooner, the match demonstrated the potential to prevent an additional 9 months or more of overpayments that would likely have occurred before SSA detected the nursing home admissions. SSA and state Medicaid officials told us that they are addressing privacy concerns that may arise from this automated match. The SSI program is the country’s largest cash assistance program for the Background poor and one of the fastest-growing federal entitlement programs. SSI benefits are available under title XVI of the Social Security Act to people who are aged, blind, or disabled and have limited income and resources. The total SSI benefit is based on the amount of income and resources the recipients report and are verified to have by SSA. The benefit consists of a basic federal payment and, in some cases, a state supplement. Page 3 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 In cases where SSI recipients are expected to be permanent residents throughout a full calendar month in Medicaid-certified medical treatment institutions2 and Medicaid pays over 50 percent of the costs of that care, the maximum SSI federal benefit is limited to $30 per month. The benefit reduction is applicable beginning with the first full month of permanent residence. If the admission is not reported promptly to SSA, a recipient may receive more than he or she is entitled to in the months following admission. In some cases, at the time of application for admission to a nursing home, a physician preliminarily determines whether the individual is expected to stay in the facility temporarily, and this information is provided to the state Medicaid agency. The agency is responsible for determining eligibility for Medicaid coverage of nursing home care. Individuals may continue to receive their full monthly benefit for up to 3 months if a physician certifies that they are expected to be institutionalized for 90 full calendar days or less and they demonstrate the need to pay expenses to maintain their home or living arrangement to which they may return upon discharge from the facility. Recipients have until the day of discharge or the 90th day of institutionalization, whichever is earlier, to provide the physician certification and statement of need. SSA does not change the benefit amount until a determination is made of whether the recipient’s nursing home stay will be temporary. SSA attempts to prevent overpayments to nursing home residents by relying on recipient self-reporting and maintaining contacts with nursing homes to obtain admissions information. In addition to attempting to prevent overpayments from occurring, SSA also uses its redetermination process, a periodic review of SSI recipients’ financial eligibility, and an annual computer match with Medicaid data provided by 28 states to HCFA to detect and stop payments that have occurred. (See app. II for more detailed information on SSA actions to prevent and detect overpayments.) SSA’s Office of Program and Integrity Reviews (OPIR), which annually Actual Extent of conducts detailed examinations of a sample of cases to determine the Overpayments Is accuracy of benefit payments, estimates that overpayments to SSI Unknown, but May recipients in nursing homes may exceed $100 million each year. OPIR identifies erroneous payments that would otherwise go undetected Exceed $100 Million because it reviews and verifies all nonmedical factors of payment eligibility for a random sample of individuals currently receiving benefits. For example, OPIR reviews include visits to institutions and recipients’ 2 These include hospitals, nursing homes, psychiatric institutions, and intermediate care facilities for the mentally retarded. For the purposes of this report, we will refer to these facilities as “nursing homes.” Page 4 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 residences to verify living arrangements. These types of in-depth examinations on a small sample of cases are in contrast to the usual procedures used in SSA field offices. There, claims representatives rely primarily on recipients to self-report changes in status because, according to SSA, it would be cost-prohibitive for field offices to conduct the same in-depth examinations. As a result, many admissions to nursing homes go undetected. In contrast to OPIR’s estimate of potential overpayments of $100 million, we found that in fiscal year 1995 SSA detected overpayments totaling $24 million to about 31,000 SSI recipients in nursing homes. Detected overpayments averaged about $800 or about 2 months of benefits; however, SSA estimates that such overpayments can continue for up to 9 months before they are detected. As shown in table 1, detected overpayments to SSI recipients in nursing homes in 1995 ranged from less than $500 to over $7,500. Our analysis of SSA’s overpayment data showed that 1,960 recipients received in excess of $2,500 in overpayments, including 386 who received more than $5,000. Table 1: Range of SSA-Detected Overpayments to Nursing Home Amount of overpayment Number of recipients Percentage of recipients Residents in Fiscal Year 1995 $1-$499 17,701 57.5 $500-$999 6,074 19.7 $1,000-$2,499 5,043 16.4 $2,500-$4,999 1,574 5.1 $5,000-$9,999 383 1.2 >$10,000 3 0.1 Total 30,778 100.0 Source: SSA’s Supplemental Security Record. SSA Is Unaware of Some We obtained nursing home admissions data for December 1996 from New SSI Recipients Recently York and Texas, two of the states that do not provide data to HCFA via the Admitted to Nursing Medicaid Statistical Information System (MSIS) and, therefore, are not included in the annual match with SSA. We determined that SSA paid Homes $515,714 in benefits to 1,699 SSI recipients in the month following their admission to nursing homes. Because these individuals were institutionalized for a full calendar month, SSI payments to them after that time are overpayments unless the recipients receive continuation of benefits due to temporary institutionalization. According to SSA records at the time of our review, these individuals were still classified as having Page 5 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 living arrangements other than institutionalization, indicating SSA may not have been aware that they were in nursing homes. Thus, SSA would have continued to erroneously pay them full benefits either indefinitely or until SSA found out about the situation. These matches indicate both that non-MSIS states have undetected overpayment situations and that they have information readily available that SSA could use to minimize its SSI overpayments to nursing home residents. We found that SSA efforts to prevent and detect overpayments to residents SSA Achieved Limited in nursing homes have had limited success. SSA has not been able to Success in Preventing effectively prevent overpayments because some SSI recipients (or their and Detecting representative payees) are not reporting changes in living arrangements as required, SSA field offices are not routinely contacting facilities to identify Overpayments SSI recipients residing in the facilities, and recent legislation requiring nursing homes to notify SSA of admissions of SSI recipients has had little effect. Likewise, additional efforts to detect overpayments have been hindered by (1) redeterminations that may be too infrequent to identify many institutionalized individuals in a timely manner and (2) an incomplete and untimely computer match with HCFA. Recipients and SSA’s first line of defense against making overpayments to nursing home Representative Payees Are residents is reports from the SSI recipients themselves. However, our Not Reporting Admissions review of SSA records indicate that some SSI recipients or their representative payees did not report changes in living arrangements as as Required required. Of the 30,778 individuals in nursing homes that SSA determined were overpaid in 1995, about 47 percent had representative payees while institutionalized. In one region we visited, SSA found that almost 75 percent of erroneous payments to individuals in nursing homes had been caused by recipients or their representative payees failing to report changes in living arrangements. Twenty-five percent of erroneous payments had resulted from field offices not following procedures when determining a recipient’s living arrangement. SSA Field Offices Not SSA field office representatives have not routinely contacted nursing homes Routinely Obtaining to solicit their cooperation in notifying SSA of admissions, as SSA policies Admissions Data From require. In October 1993, SSA established policies requiring all field offices to work closely with the staffs of nursing homes to facilitate the flow of Facilities information regarding the admission to nursing homes of SSI recipients. We found that most field offices we contacted had not established working Page 6 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 relationships with the facilities, and in some cases they were not even aware of the facilities in their area of responsibility. According to the field office personnel we interviewed, some nursing homes have routinely notified SSA field offices of SSI recipient admissions; however, field offices could not always account for the notifications because some do not maintain a log or have standard procedures for following up on notifications. We found that neither SSA headquarters staff nor field office representatives routinely monitor facility notifications to ensure follow-up. Each field office manager establishes the office’s work priorities, and in the offices we visited we found that the priority placed on following up on facility notifications varied. We were told that facility notifications are given much lower priority than work responsibilities that are monitored, such as claims processing. In addition, both SSA field office representatives and regional office representatives stated that some notifications would not have been processed in a timely manner, or in some cases not at all, because they had been misplaced or lost. We were unable to quantify the number of untimely or unprocessed facility notifications because no records of notifications have been maintained. Effect of Recent The effect that recent legislation3 requiring nursing homes to report Legislation in Preventing admissions to SSA has had in preventing SSI overpayments to nursing home Overpayments Is residents is difficult to determine; SSA does not have a uniform system for following up on admission notifications from nursing homes or for Questionable monitoring compliance with the law. However, our analysis of overpayment data and interviews with SSA headquarters and field office officials indicate that the legislation has had little effect in preventing overpayments. For example, the amount of detected overpayments to SSI recipients in nursing homes has grown by 12.3 percent since the October 1995 effective date of the legislation, increasing from $24 million in fiscal year 1995 to $27 million in fiscal year 1996. In instances where facilities had reported admissions of SSI recipients, the majority of the field offices we visited had no system for documenting the receipt or disposition of the admissions referrals to SSA. As a result, SSA has no data on which to determine whether nursing homes are complying with the law, and there is no assurance that field office claims representatives are following up on all admissions notifications. 3 The Social Security Domestic Employment Reform Act of 1994 (P.L. 103-387, sec. 6, Oct. 22, 1994). Page 7 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 The effectiveness of the law has likely been diminished by limited SSA efforts to carry out the reporting requirement and the lack of a penalty for nonreporting by nursing home administrators. SSA has not developed any regulations establishing a uniform mechanism for nursing homes to report admissions or revised its agency policies on coordinating with institutions since enactment of the law. Instead, SSA efforts primarily focus on having its 1,300 field offices maintain contacts and solicit information from the over 23,000 U.S. nursing homes. SSA has publicized the nursing home reporting requirement and HCFA has issued notices to nursing home administrators informing them of their reporting responsibilities. The legislation, however, does not have a penalty for nonreporting by nursing home administrators, and the amount of reimbursement nursing homes receive for treating Medicaid patients is not affected by reporting or not reporting; therefore, SSA must rely on voluntary compliance by nursing homes. In 1995, SSA requested that HCFA develop and implement procedures for monitoring compliance with the reporting requirement. To date, neither SSA nor HCFA has developed such a system. Often Redeterminations SSA uses its redetermination process to verify that recipients remain Have Not Identified financially eligible for SSI payments and are receiving the correct amounts. Institutionalized However, because of resource constraints, SSA reviews the eligibility of most recipients only once every 6 years. SSA records indicated that in 1995, Individuals in a Timely 4,792 of the 30,778 overpaid individuals had redeterminations while they Manner were in nursing homes. We found that 3,099 individuals each had one redetermination, 352 had two, and 60 had three or more. According to SSA records, 364 of these redeterminations involved face-to-face contact between an SSA field office employee and the recipient or the representative payee. Because of the infrequency of some redeterminations, SSA cannot rely on this process to routinely and effectively identify in a timely manner overpayments due to nursing home residency. Current Computer Match Although SSA’s computer match with state Medicaid data from HCFA results With HCFA Is Incomplete, in about $4 million in program savings each year, it only contains data Not Timely, and Results in from 28 states, identifies overpayments only after they have continued for a lengthy period, and places an unnecessary work burden on field offices. Unnecessary Field Office SSA does not independently obtain Medicaid nursing home admissions data Work from states not participating in MSIS. Consequently, admissions of SSI recipients to nursing homes in the remaining states, unless self-reported, are likely to go undetected for long periods. Page 8 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 In addition, SSA’s computer match with HCFA is not timely. For example, in June 1995 HCFA matched data on possible nursing home admissions occurring between April 1, 1994, and March 31, 1995. The resulting match cases were sent to SSA field offices in October 1995 to be included in their annual workload.4 Based on our review of a sample of cases identified using the match, we found that an average of 14 months elapsed from the time an individual was admitted to a nursing home until SSA headquarters notified the field office to review the person’s case. Furthermore, it took field office representatives an additional 4 months before they reviewed the cases and made changes, if necessary, to the benefit amount. The match also incorrectly identifies many individuals as having changed their living arrangement, therefore placing an unnecessary and unproductive work burden on field offices. Field office officials told us that many of the match cases they review result in no changes in recipients’ living arrangements or benefit amounts. We analyzed a sample of 1996 MSIS match cases sent to field offices at the beginning of the fiscal year and found that as of the close of the year, SSA field offices had not completed reviews of 141 (28 percent) of the 503 cases in our sample. About 28 percent of the completed cases were erroneously selected for review because the individuals were temporarily institutionalized and had been granted a continuation of full benefits. Another 26 percent of the completed cases identified individuals who did not require a change in benefit amount. SSA officials are aware of the deficiencies in the match selection criteria that result in temporarily institutionalized individuals or those in nursing homes for less than a full calendar month being selected for review. However, they have yet to change the criteria to only identify individuals institutionalized for a full calendar month who have not been granted a continuation of full benefits because of temporary institutionalization. Obtaining MMIS data on nursing home admissions directly from states and Automated Interface conducting an automated interface in accordance with laws and standards Could Help Prevent governing computer matching, privacy, and security could provide SSA Overpayments with the opportunity to prevent or detect erroneous payments more quickly, without the use of SSA field office personnel, and could result in program savings and reduced administrative costs. The cost and ease of states making MMIS data directly available to SSA electronically will vary depending upon the level of automation in each state. However, we have 4 Field office workloads consist of processing initial claims and completing postentitlement actions, such as redeterminations, benefit recomputations, and address changes. Page 9 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 discussed this with selected state Medicaid, HCFA, and SSA officials, who agreed that such a data exchange would be both practical and desirable. Currently, for those who do not self-report their nursing home admission, our review of a sample of 1996 cases found that it takes an average of 14 months from the time an individual enters a facility until the overpayment is detected using the annual computer match with HCFA data that are only available from 28 states, and takes an additional 4 months for the benefit amount to be reduced by SSA. By electronically obtaining nursing home admissions data directly from state Medicaid agencies SSA could prevent overpayments or detect them much sooner—1 to 3 months after they begin—for all 50 states. For example, an automated interface could consist of matching SSI payment data with monthly nursing home admissions data from state Medicaid Management Information Systems (MMIS) obtained using existing telecommunications lines.5 According to SSA and HCFA officials, SSA could obtain the data directly from states by modifying existing state data exchange agreements6 to make SSA the recipient agency of state MMIS data on nursing home admissions. A way to efficiently use state MMIS data on nursing home admissions would be through an automated system. An individual identified in the data as residing in a facility for a full calendar month and subject to a benefit reduction could automatically be sent a notice generated by computer explaining the detection of a potential overpayment situation, the potential revised payment amount, criteria for receiving the continuation of benefits if the recipient is temporarily institutionalized, and the process for appealing the benefit reduction. Following a reasonable response time period to provide due process, the computer could, in appropriate cases, automatically adjust the SSI benefit to the correct amount for the appropriate months’ payments. Field office claims representatives would not have to review the case or manually input changes to the recipient’s payment file in order to correct the benefit amount. This would minimize the time period over which the overpayment occurs, thereby saving program dollars, and would reduce field office time devoted to this activity, freeing it for other purposes. In those cases in which individuals requested, and were determined to be eligible for continuation of benefits 5 These lines, known as the File Transfer Management System (FTMS), already exists between SSA headquarters and the states. SSA installed FTMS so that it could transmit SSA data on its clients to every state. State agencies are required by the 1984 Deficit Reduction Act to use this information to better identify those who are not eligible for public assistance or who are receiving incorrect benefits. 6 At least monthly, SSA makes available to all states SSI eligibility and payment information to assist them in administering SSI state supplemental payments and other federally funded programs such as Medicaid and food stamps. Page 10 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 due to temporary institutionalization, SSA could manually input changes to the recipients’ files, overriding the automatic benefit reduction and thereby continuing uninterrupted full benefit payments. Such an automated system would require testing on SSA’s part to ensure the reliability of state data for making automatic payment reduction decisions and to minimize the risk of inappropriate reductions. Some states have already demonstrated their capability to share data with SSA to prevent and detect overpayments. Currently, the SSI payment file is matched with data from the 28 states participating in MSIS. Furthermore, in 1994, SSA began establishing direct connections between its field offices and states that had automated databases that could be easily linked to SSA’s computer system. Currently, 15 states have entered into agreements allowing SSA field offices to directly access certain state databases containing information SSA can use to verify recipients’ reported income and other factors of financial eligibility. State Medicaid and SSA field office officials we interviewed said that SSA’s routinely obtaining nursing home admission data directly from states could provide the best opportunity to prevent or detect overpayments to SSI recipients in nursing homes. Three of the states we visited do not participate in MSIS, but the Medicaid officials in these states said that they can make nursing home admissions data available to SSA electronically. Tennessee has already made various state databases available for use by SSA field offices. New York and Texas are negotiating with SSA to establish pilot projects granting SSA access to certain state databases. Pilot Match With To help demonstrate the feasibility of SSA’s directly obtaining state Tennessee Demonstrates Medicaid nursing homes’ admissions data and more quickly detecting Feasibility and overpayments, we coordinated a joint effort between SSA and the Tennessee Department of Health Services to identify SSI recipients in Effectiveness of Directly nursing homes. On the basis of data we obtained from Tennessee and Obtaining State Data provided to SSA, SSA detected $31,000 in overpayments to 40 SSI recipients from February to July 1996. These recipients had not previously been identified by SSA as residing in nursing homes. The overpayment amounts ranged from less than $10 to almost $1,800, and the average overpayment was $775. In addition to detecting overpayments, the pilot allowed SSA to prevent at least 9 months of additional overpayments to individuals that would likely have gone undetected and that SSA would not have been likely to collect. The pilot match in Tennessee, a state that accounts for about 3 percent of all SSI recipients, demonstrates that by directly obtaining state Page 11 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 data, SSA could not only more quickly identify SSI recipients in nursing MMIS homes, but also prevent future overpayments. SSA Currently Conducts SSA already conducts routine automated computer interfaces with the Automated Interfaces With Department of Veterans Affairs, the Office of Personnel Management, and Some Federal Agencies the Railroad Retirement Board that result in automatic benefit reductions without SSA field office involvement. In each of these interfaces, SSR and files from the source agencies are matched. When the match determines that an individual on both the SSR and the source agency file are the same person and there has been a change in income affecting the SSI benefit, the system automatically adjusts the benefit and generates a notice to the recipient about the revised payment amount. Before any benefit reduction occurs, a recipient has 10 days after receiving notification of the benefit change to file an appeal and continue receiving unreduced benefits, otherwise the benefit reduction occurs automatically, without a field office representative reviewing the case. SSA estimates that these interfaces result in about $41 million in program savings each year. SSA and States Have Taken Certain privacy and security concerns may arise when data are exchanged Actions That Address electronically between agencies. These concerns center on ensuring that Privacy and Security personal information that an individual provides to one government agency is protected from being disclosed to other agencies that do not Concerns Raised by have a legal right to it. Granting SSA direct access to state Medicaid data Electronic Data Exchanges will not violate the privacy rights of individuals who provide this information because SSA will simply electronically obtain information to which it already has a legal right. SSA already obtains this information from 28 states using HCFA’s MSIS data. As part of its procedures to determine whether to implement electronic data exchanges with state agencies, SSA assesses the costs, benefits, and security risks of conducting such exchanges. According to SSA officials, an automated computer interface between SSA and state MMIS would be subject to the same procedures and feasibility testing prior to nationwide implementation. Although we did not evaluate the effectiveness of SSA’s or the states’ security procedures, SSA and state officials told us that these procedures will be stringent enough so that SSA can obtain state data electronically and conduct automated interfaces without compromising confidentiality. SSA and states have taken steps specified in federal security standards that, we were told, would ensure the confidentiality and security of data exchanged electronically. These include instituting written Page 12 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 agreements between SSA and state agencies regarding how the data will be used and using computer lines dedicated solely to the transmission of data between government agencies. States Can Provide Data to In the interim period, while states are preparing to make the necessary SSA in Other Formats nursing home admissions data directly available to SSA electronically, they While Preparing to Make It could provide SSA with tapes or paper listings of admissions to nursing homes. SSA could use this information to detect overpayments sooner than Available Electronically it can using data from the current annual computer match. In the states that we contacted during our review, state officials told us that providing this information to SSA routinely would require only minimal computer programming to format the data for SSA’s use. Two states we visited, New York and Texas, provided this information to us on computer tapes so that we could identify previously undetected potential overpayments to nursing home residents in those states. Generally, SSI benefits are supposed to be reduced for those individuals in Conclusions nursing homes for more than a full calendar month when Medicaid is paying the cost of care. For many years, however, SSA has lacked an effective process to prevent SSI recipients in nursing homes from receiving overpayments. It has relied primarily on (1) inadequate self-reporting by recipients, (2) inconsistent and irregular field office contacts with nursing homes, (3) untimely redeterminations, and (4) an incomplete computer match with Medicaid data from HCFA that does not result in timely identification of nursing home admissions. Neither SSA’s efforts nor the recent legislation requiring nursing homes to report admission of SSI recipients to SSA has been fully effective. As a result, SSA continues to overpay millions of dollars to thousands of recipients in nursing homes. Given SSA’s experience that only about 15 percent of outstanding overpayments to SSI recipients are collected, it is important that SSA detect these overpayments as soon as possible and prevent future overpayments. HCFA has the authority to require states to make available to SSA MMIS information on nursing home admissions. States would be paid 90 percent of the costs of developing the necessary capabilities and 75 percent of the operating costs. Our efforts with three states have shown that directly obtaining state MMIS data could help SSA prevent or more quickly detect overpayments and simultaneously reduce the SSA field office workload. In the interim, while such arrangements are being made, states could provide tapes or hard copies of these data to SSA to help control its payments, with Page 13 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 minimal effort or cost to the states. Preventing or detecting erroneous payments more quickly and decreasing SSA’s reliance on recipients and nursing homes to report changes in circumstances that affect eligibility would bolster the integrity of the SSI program by helping ensure that clients are receiving only those benefits to which they are entitled and would save both program and administrative costs. To prevent overpayments to SSI recipients in nursing homes or detect them Recommendations sooner, we recommend that the Secretary of Health and Human Services direct the Administrator of HCFA to require states, as part of MMIS systems requirements, to include information on nursing home admissions as standard data elements in their MMIS and make these data elements available to SSA electronically, in accordance with the laws and standards governing computer matching, privacy, and security. We also recommend that the Commissioner of SSA take the following actions: • Establish agreements with the states to routinely obtain state MMIS data on nursing home admissions electronically, as soon as feasible. • Establish interim agreements with state Medicaid agencies, while states adapt their systems to make this information available to SSA electronically, to obtain computer tapes or paper listings of admissions to nursing homes and use this information to identify overpayment situations and begin recovery actions and payment reductions. • Determine the reliability of state MMIS data for purposes of supporting automatic benefit reductions for those SSI recipients identified as residing in nursing homes for a full month who are not eligible for continuation of full benefits due to temporary institutionalization and, if the data are reliable, implement a system for automatic benefit reduction. In commenting on a draft of this report, SSA agreed that improvements can Agency Comments be made in obtaining and processing nursing home admissions data on SSI and Our Evaluation recipients (see app. III). SSA noted that it has been working on solutions to this problem, including obtaining MSIS data, which are submitted voluntarily by 28 states, from HCFA. Acknowledging that MSIS data do not cover all admissions, SSA is working with HCFA to determine if HCFA’s new system, the Resident Assessment Instrument System (RAIS), can be used to identify SSI recipients in nursing homes in a more timely manner. The primary purpose of collecting the assessment information is to help Page 14 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 nursing home staff plan and evaluate the care they provide to residents. SSA said that should RAIS not prove feasible, it would then consider the recommendations in our report. We have no objection to SSA’s use of RAIS, if in fact it is the most effective and efficient way of reducing SSI overpayments to nursing home residents. However, while RAIS may provide SSA with the data it needs to more quickly identify SSI recipients in nursing homes, we have several significant concerns in this regard. First, not all states participate in RAIS. While HCFA has instructed all Medicare- and Medicaid-certified nursing facilities to complete the assessment on all residents upon admission, currently HCFA does not require that the assessment results be submitted to it or any other entity. According to HCFA officials, HCFA will require all states to participate in RAIS as part of the President’s fiscal year 1998 budget proposal to create a separate prospective payment system for nursing homes. HCFA officials told us that they do not know whether this proposal will be adopted. However, contrary to SSA’s comments on our report, HCFA already has the authority to require all states to make nursing home admissions data available to SSA using state MMIS, as we suggested in this report. Second, we are concerned about the reliability of the RAIS data in identifying whether Medicaid is paying for the nursing home care, which is a key factor in determining whether a recipient’s SSI benefit amount is affected by residency in a nursing home. According to HCFA officials, facilities’ nursing staffs would most likely be completing the assessments and may not be familiar with the source of payment for the residents. If the nurses do not accurately complete the source of payment information on the assessment forms, SSA would be matching its file with erroneous payment data from RAIS. The MMIS information we recommend that SSA use is treatment authorization or paid claims data already approved by the state Medicaid agency. A third issue concerning RAIS is timeliness. RAIS is still being piloted and, according to HCFA officials, may not be fully operational in all states at least until summer 1998, which would allow the current level of related SSI overpayments to continue for over a year. SSA stated in its comments on this report that MMIS data would be available only quarterly. However, MMIS data can be provided monthly or even more frequently if sent from the states directly to SSA. Consequently, we continue to believe that electronically obtaining nursing home admission data directly from states provides SSA with the best opportunity to detect nursing home admissions Page 15 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 more quickly than its current efforts and also to prevent additional overpayments. In the interim, while SSA determines whether RAIS is the best solution for identifying SSI recipients in nursing homes, we believe that obtaining MMIS data directly from states each month through computer tapes or paper listings is preferable to the delays experienced under SSA’s current procedures. The states we contacted said that they could format the data for SSA’s use and provide data to SSA on a monthly basis. Furthermore, two states we visited demonstrated that nursing home admissions data could be made available by providing us with computer tapes that allowed us to identify previously undetected potential overpayments. Finally, SSA said that it is highly questionable that a system for automatic benefit reduction can be implemented because of the in-depth information required to determine whether a recipient is eligible for continued benefits due to temporary institutionalization. As we stated in the report, the benefit reduction would occur only after automated notices were sent to recipients asking if they resided in the nursing home for more than a full calendar month and the appropriate due process procedures were followed. For those recipients requesting a continuation of benefits, the automated reduction would be suspended, and SSA would then follow its current procedures for determining eligibility for continued benefits. Only the benefits of those recipients remaining in nursing homes and not requesting a continuation of benefits would automatically be reduced without field office review. In commenting on a draft of this report, HHS noted HCFA’s willingness to provide the MSIS data it now receives from the states to SSA on a quarterly rather than annual basis. (See app. IV.) This suggestion fails to recognize the underlying reason for our specific recommendation. Having HCFA involved in the pass-through of information between the states and SSA causes both an unnecessary expense for HCFA and a timeliness delay for SSA. The SSI program needs access to nursing home admissions data as quickly as possible after the actual admissions to minimize the overpayments it makes. Consequently, we continue to believe that the most effective data transfer is directly between the states and SSA on as frequent a basis as possible. HHS also commented about the need to offset the costs associated with our recommendation for each state and for HCFA against the benefits to be derived by the SSI program. We agree. As stated in this report, we were told Page 16 GAO/HEHS-97-62 Overpayments to Nursing Home Residents B-271862 by both officials of the states we visited and knowledgeable officials in HCFA that only minimal computer programming would be required to format the data for SSA’s use. It should also be noted that the federal government would pay 90 percent of the developmental costs and 75 percent of the operating costs for operating such a data exchange system, thereby further reducing the burden of this change on the states. The minimal costs associated with developing and operating such a system, including those incurred by SSA, should be compared with the tens of millions of dollars in reduced overpayments to the SSI program that our work indicates would accrue. Moreover, if our recommendations are adopted, HCFA will no longer be asked to provide MSIS data to SSA, thereby saving HCFA the costs it currently incurs. Finally, HHS suggested that we also investigate using the RAIS database as another alternative to our recommendation. As stated in our response to SSA’s comments on this report, it is unknown when all states will participate in contributing data on nursing home admissions to RAIS, and the reliability of this system for identifying whether Medicaid is paying for the nursing home care is questionable. This information is paramount to determining whether an SSI overpayment has been made. SSA and HHS also made other technical comments, which we incorporated throughout the report as appropriate. We are sending copies of this report to relevant congressional committees; the Director, Office of Management and Budget; the Commissioner, SSA; the Secretary of Health and Human Services; and other interested parties. Copies will also be made available to others on request. If you or your staff have any questions concerning this report, please call me on (202) 512-7215. Other GAO contacts and staff acknowledgments are listed in appendix V. Jane L. Ross Director, Income Security Issues Page 17 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Contents Letter 1 Appendix I 20 Objectives, Scope, and Methodology Appendix II 22 SSA Actions to Prevent and Detect Overpayments Appendix III 25 Comments From the Social Security Administration Appendix IV 29 Comments From the Department of Health and Human Services Appendix V 33 GAO Contacts and Acknowledgments Table Table 1: Range of SSA-Detected Overpayments to Nursing Home 5 Residents in Fiscal Year 1995 Page 18 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Contents Abbreviations FTMS File Transfer Management System HCFA Health Care Financing Administration HHS Department of Health and Human Services MMIS Medicaid Management Information System MSIS Medicaid Statistical Information System OPIR Office of Program and Integrity Reviews RAIS Resident Assessment Instrument System SSA Social Security Administration SSI Supplemental Security Income program SSR Supplemental Security Record Page 19 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix I Objectives, Scope, and Methodology To determine the extent of overpayments to individuals because they resided in nursing homes, we reviewed the Supplemental Security Record, the Social Security Administration’s payment record for the Supplemental Security Income program. We obtained fiscal year 1995 and 1996 data and identified those individuals SSA determined had been paid erroneously because of institutionalization in nursing homes. We analyzed the payment record and developed aggregate information on overpaid recipients. To determine the success or failure of SSA’s actions in preventing and detecting overpayments, we interviewed SSA headquarters officials in Baltimore, Maryland; 4 regional offices; and 13 field offices. In addition to these discussions, we analyzed the SSR to determine the effectiveness of SSA’s match with the Health Care Financing Administration. Of the 5,907 fiscal year 1996 Medicaid Statistical Information System match cases sent to the field offices for review, we identified the 503 cases that appeared in SSA’s 10-percent sample data files. These files are random samples of all cases in a given month on the SSR. SSA field offices had completed reviews of 362 of the 503 cases as of the end of the fiscal year. We obtained and analyzed payment history data for the completed cases to determine the outcome of SSA’s matching effort. Finally, to determine methods by which SSA can better prevent overpayments, we interviewed state Medicaid officials in five states to obtain information about Medicaid data systems to determine the feasibility of conducting an automated interface with SSA. We selected Tennessee because of its advancements in providing electronic access to the state’s information systems. We selected the other four states (California, Florida, New York, and Texas) because they have the four largest SSI populations, accounting for almost 40 percent of the total SSI population. We obtained Medicaid nursing home admissions data for December 1996 from New York and Texas and matched it against the SSI payment file to identify SSI recipients admitted in that month to nursing homes in these states and the resulting potential overpayments. Moreover, we coordinated a pilot between SSA and the Tennessee Department of Health Services to further document the feasibility of SSA’s directly obtaining state Medicaid data. Tennessee officials provided data on all individuals in Tennessee approved for admission to nursing homes in February 1996 and paid claims data from February through July 1996. We used these data to determine the length of nursing home stays. We provided this information to SSA and had SSA systems analysts develop a computer program to compare Tennessee’s admissions and paid claims Page 20 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix I Objectives, Scope, and Methodology data with SSR data to identify SSI recipients who had received overpayments because they resided in nursing homes. Since SSR and the state Medicaid information systems are subject to periodic quality assurance reviews, we did not independently examine the computer system controls for them. Except for this limitation, we conducted our review between April 1996 and April 1997 in accordance with generally accepted government auditing standards. Page 21 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix II SSA Actions to Prevent and Detect Overpayments SSA requires that recipients report entrance into nursing homes as quickly SSA Relies on as possible. However, because many recipients do not report this Reporting by information, SSA also establishes working relationships with nursing Recipients and homes to obtain admission and discharge information on the SSI population. Furthermore, legislation was enacted, effective in Nursing Homes to October 1995, to require that nursing homes provide this information to Prevent SSA. Overpayments At the time of application for benefits, claims representatives in SSA field offices inform SSI recipients that they are required to report information that may affect their eligibility or payment amounts. If the recipient has a representative payee, the payee is responsible for reporting such information to SSA. Significant events to be reported include a change in income, resources, marital status, or living arrangements, such as admission to or discharge from a nursing home. Failure to report such changes can result in monetary penalties of up to $100 per event. Although SSI recipients are the primary source for reporting changes in their living arrangements, SSA recognizes that recipients entering nursing homes may not always report their admissions on a timely basis, if at all. Therefore, SSA attempts to obtain admission and discharge information directly from nursing homes. In October 1993, SSA instructed its field offices to work closely with nursing homes to facilitate the flow of information regarding the admission of SSI recipients. Field offices are to maintain ongoing contact with all appropriate institutions in their service areas, use regular visits as a means to encourage cooperation, and establish procedures for institutions to make timely reports on events that affect SSI recipients’ eligibility and benefit amount. Also, the Social Security Act, as recently amended, states that the Commissioner of SSA must require each administrator of a nursing home to report within 2 weeks any admission occurring on or after October 1, 1995, of any eligible individual or eligible spouse receiving SSI benefits. The legislation was designed to prevent overpayments to SSI recipients who failed to report their admission to nursing homes. To comply with the legislation, SSA and HCFA notified nursing home administrators of the reporting requirement. The law, however, did not contain a penalty for nonreporting by nursing home administrators. In addition to relying on recipient self-reporting and contacts with nursing SSA Actions to Detect homes to prevent overpayments, SSA uses its redetermination process and Overpayments Once an annual computer match to detect such payments once they have They Have Been Made occurred. A redetermination is a review of financial eligibility factors to Page 22 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix II SSA Actions to Prevent and Detect Overpayments ensure that recipients are still eligible for SSI payments and are receiving the correct amount. A redetermination addresses financial eligibility factors such as income, resources, and living arrangements and can be conducted by mail, telephone, or face-to-face interview. Given its limited resources, SSA conducts redeterminations on over two-thirds of the SSI population receiving benefits approximately once every 6 years but may conduct them more frequently if it determines that changes in eligibility or erroneous payments are likely. Recipients with a history of recent earnings are likely to be redetermined more often than recipients, such as institutionalized individuals, who generally do not experience fluctuations in their income or resources. The redetermination procedure includes a question for recipients about whether the recipient spent a full calendar month in a hospital, nursing home, other institution, or any place other than the recipient’s normal residence. With this question, SSA hopes to identify situations where overpayments to recipients may have occurred. SSA has conducted an annual computer match with HCFA since 1992 to identify SSI recipients in nursing homes. Currently, SSA sends a file to HCFA containing identifying information on all SSI recipients residing in the 28 states that provide data to HCFA via the Medicaid Statistical Information System (MSIS). The MSIS file contains Medicaid usage data submitted by states on a voluntary basis. HCFA matches the MSIS nursing home admissions data with the SSA file to identify SSI recipients who resided in nursing homes during the period covered by the match and sends a file to SSA with this information. After reviewing the match results, SSA deletes cases in which individuals (1) have self-reported their admission to SSA, (2) are not in current pay status, or (3) are deceased. SSA then distributes information on potential overpayment situations to the appropriate field offices during October of each year. According to SSA procedures, field office representatives are to contact the recipients or nursing homes during the balance of the fiscal year to determine if overpayments have been made. SSA does not begin overpayment recovery efforts until recipients are discharged from facilities and are eligible to receive their full SSI benefit. Although SSA uses the MSIS data from 28 states to help detect overpayments, Medicaid data from all states exist and could be used by SSA more frequently than once a year. Generally, nursing homes report admissions of individuals and file claims for reimbursement with state Medicaid agencies in a timely manner to ensure rapid payments from Medicaid to them. States use automated systems, known as Medicaid Management Information Systems (MMIS), to process claims and to capture Page 23 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix II SSA Actions to Prevent and Detect Overpayments and report data needed by HCFA and the states to manage the Medicaid program. The Secretary of Health and Human Services has broad authority to require states to report data that HCFA needs to administer the Medicaid program. State MMIS must meet performance standards as well as system and compatibility requirements established by HCFA. HCFA requires each state’s MMIS to include 122 standard data elements and pays 90 percent of the development costs and 75 percent of the operating costs of these systems. Page 24 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix III Comments From the Social Security Administration Page 25 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix III Comments From the Social Security Administration Page 26 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix III Comments From the Social Security Administration Page 27 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix III Comments From the Social Security Administration Page 28 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix IV Comments From the Department of Health and Human Services Page 29 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix IV Comments From the Department of Health and Human Services Page 30 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix IV Comments From the Department of Health and Human Services Page 31 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix IV Comments From the Department of Health and Human Services Now on p. 9. Now on p. 4. Now on p. 9. Page 32 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Appendix V GAO Contacts and Acknowledgments Roland H. Miller III, Assistant Director, (202) 512-7246 GAO Contacts George A. Scott, Evaluator-in-Charge, (202) 512-5932 In addition to those named above, the following individuals also made Acknowledgments important contributions to this report: Mary Ellen Fleischman, Evaluator; James C. Lawson, Evaluator; Graham D. Rawsthorn, Evaluator; John G. Smale, Jr., Senior Social Science Analyst; and James P. Wright, Assistant Director (Study Design and Analysis). (106809) Page 33 GAO/HEHS-97-62 Overpayments to Nursing Home Residents Ordering Information The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. Orders by mail: U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 or visit: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists. For information on how to access GAO reports on the INTERNET, send an e-mail message with "info" in the body to: email@example.com or visit GAO’s World Wide Web Home Page at: http://www.gao.gov PRINTED ON RECYCLED PAPER United States Bulk Rate General Accounting Office Postage & Fees Paid Washington, D.C. 20548-0001 GAO Permit No. G100 Official Business Penalty for Private Use $300 Address Correction Requested
Supplemental Security Income: Timely Data Could Prevent Millions in Overpayments to Nursing Home Residents
Published by the Government Accountability Office on 1997-06-03.
Below is a raw (and likely hideous) rendition of the original report. (PDF)