United States General Accounting Office GAO Report to Congressional Committees May 1999 MEDICARE SUBVENTION DEMONSTRATION DOD Data Limitations May Require Adjustments and Raise Broader Concerns GAO/HEHS-99-39 United States GAO General Accounting Office Washington, D.C. 20548 Health, Education, and Human Services Division B-278140 May 28, 1999 Congressional Committees The Balanced Budget Act of 1997 (BBA) authorized a 3-year test, called Medicare subvention, allowing Medicare-eligible military retirees, their dependents, and survivors to enroll in a new Department of Defense (DOD) health maintenance organization (HMO). The demonstration’s stated goal is to implement an alternative for delivering accessible and quality care to Medicare-eligible military beneficiaries, while not increasing the cost to either DOD or Medicare. Currently, care for these beneficiaries at military treatment facilities (MTF) is provided on a space-available basis that lacks the continuity often important to older retirees. Under this demonstration, the Medicare Trust Funds will pay DOD for health care provided to eligible retirees at six sites. DOD will provide enrollees the full range of Medicare-covered services as well as some additional services. In principle, beneficiaries, DOD, and Medicare could all gain under subvention. Beneficiaries who choose DOD’s plan can use their Medicare benefit to receive care at an MTF. Under subvention, Medicare’s payment for enrollees could be less than what it pays private plans serving other Medicare beneficiaries, and DOD could gain additional funds and use excess capacity where it exists. The BBA required that, before Medicare reimburses DOD under the demonstration, the test sites spend the amount they would have spent without the demonstration on Medicare-eligible retirees’ care.1 DOD already receives money for its care of retirees aged 65 and over as part of its annual appropriation. Since DOD does not have an accounting system that can measure the cost of care provided to individuals, DOD developed, and the Health Care Financing Administration (HCFA), within the Department of Health and Human Services, agreed with, a complex method to estimate this “level of effort” (LOE), or baseline. It is important that LOE be correctly calculated. If LOE is underestimated, Medicare may overpay; if LOE is overstated, Medicare may underpay, which could cause DOD to further reduce space-available care or shift resources from other programs or beneficiary groups to pay for demonstration enrollees’ care. Using 1996 data, DOD currently estimates its LOE for the six sites to be $172 million.2 To further protect the Trust Funds, the BBA caps payments to DOD at 1 More precisely, the requirement is in the Social Security Act, as amended by the BBA. (Section 4015 of the BBA, P.L. 105-33, 111 Stat. 251, 337, added section 1896 to the Social Security Act. This section authorizes the subvention demonstration. See 42 U.S.C. 1395ggg.) 2 App. I describes in more detail the process used to determine Medicare payments. Page 1 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 $50 million in the demonstration’s first year, $60 million in the second year, and $65 million in the third year. The BBA also directed GAO to report annually on the demonstration’s effect on Medicare costs.3 Because the demonstration began delivering care at its first site in September 1998 and was not fully implemented at all sites until January 1999, there is not yet sufficient evidence to assess subvention’s cost to Medicare. Consequently, this first report to your committees focuses on the sufficiency of DOD’s data systems for (1) determining DOD’s historical LOE and Medicare payments and (2) managing the demonstration and assessing its cost effects. In conducting our evaluation, we reviewed not only DOD’s method for measuring LOE and capturing DOD health care costs but also source data from key DOD information and accounting systems used to calculate LOE and manage the military health care system in general. We conducted our review in accordance with generally accepted government auditing standards. (Addressees are listed at the end of this letter. App. II describes the scope and methodology of our work in more detail.) Portions of DOD’s baseline costs may be understated, which could lead to Results in Brief Medicare overpayments if not adjusted. This results from data inaccuracies in areas of DOD’s medical cost accounting system such as pay and prescription drugs. Our findings show that the DOD cost system problems we and others have reported on over the years continue to affect the DOD health care activities that rely on these systems. At the root of the problem is the long-standing lack of DOD and services’ oversight as well as a lack of incentives to ensure the data’s accuracy, timeliness, and completeness. DOD officials told us that DOD is committed to making the adjustments necessary to ensure Medicare does not overpay DOD. Data problems also make the subvention demonstration more difficult to manage at both the national and local levels. For example, DOD managers do not have sufficiently accurate or timely data to know whether Medicare capitated payments will cover DOD’s costs to provide the full range of health care to beneficiaries or to determine whether it is more cost-effective to deliver care in DOD facilities or purchase it from network providers. Timely and accurate tracking of cost and utilization data is critical to these decisions, as is the case in other managed care organizations. 3 We are to report on a number of other issues, including the demonstration’s impact on access, quality, and military readiness, as well as DOD’s management of the demonstration and compliance with Medicare regulations. These issues will be the subjects of future reports. Page 2 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 Acting on the problems we identified, DOD officials developed a management improvement plan to begin addressing baseline and systemic data weaknesses, and HCFA plans to hire a contractor to review DOD’s data and methodology. In their reviews, these agencies may need to reestimate the baseline using more reliable data or consider alternate ways to determine the baseline. Because DOD uses its cost accounting systems for many other health care management purposes beyond the demonstration’s needs, such as resource allocation and “make-versus-buy” decisions, we believe DOD needs to dedicate sufficient management attention and effort to ensure data reliability and accuracy. Recently, DOD established a health care data quality task force to begin addressing the broader system causes of the data problems that we and others have continued to identify. We make several recommendations in this report concerning these matters. Currently, about 1.3 million retired military personnel and their Background dependents and survivors who reside in the United States are age 65 or older. This number is expected to increase to about 1.6 million by 2004. By contrast, the number of active duty personnel and their dependents is projected to remain constant. Of the 1.3 million dual eligibles (that is, eligible for both military health care and Medicare), about half live within 40 miles of an MTF. This 40-mile radius is a rule of thumb for defining such facilities’ “catchment” (or service) areas. Retirees are eligible for a broad range of health care services under TRICARE Prime,4 DOD’s HMO program, until they turn 65, when they become eligible for Medicare. Once they turn 65, retirees lose their eligibility for TRICARE Prime. They continue to qualify for inpatient and outpatient care in MTFs, but only on a space-available basis. Limited space and resources, coupled with the priority given to active duty personnel and other beneficiaries who are under age 65, mean that military retirees aged 65 and over often do not get appointments and other services at an MTF when they need them (although they may continue to get prescription drugs from MTFs). Most military retirees who are 65 and over are eligible for Medicare, a federal program administered by HCFA that covers health care expenses of the elderly, some disabled people, and people with end-stage kidney 4 As an employer, DOD established its TRICARE program to provide comprehensive health care to active duty personnel, their dependents, and military retirees. TRICARE beneficiaries may get care at MTFs as well as from civilian providers in the local community. Page 3 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 disease. Medicare part A covers inpatient hospital, skilled nursing facility, and hospice care; Medicare part B covers physician and other outpatient services for beneficiaries choosing to pay a monthly premium. Original, or traditional, fee-for-service Medicare has two distinctive features: it allows the patient to choose his or her physician, and it reimburses beneficiaries’ claims for hospital, physician, and other care on a fee-for-service basis. Beneficiaries who receive care are responsible for part of the charges—for example, 20 percent of the Medicare fee schedule amount for physician services, or the $768 deductible for hospital care. As an alternative to fee-for-service Medicare, beneficiaries may choose the Medicare+Choice option, which permits them to enroll in private Medicare HMOs and other private health plans. These plans provide all standard Medicare benefits. Beneficiaries in these plans, like beneficiaries in original Medicare, must pay the program’s monthly premium for part B coverage. Medicare+Choice plans also may offer additional benefits, such as prescription drug coverage, and may waive cost-sharing required by original Medicare. For these additional benefits, plans may charge an extra premium, though many do not. Medicare pays a capitated rate (a fixed amount each month per enrollee) to Medicare+Choice plans, and the plans bear the financial risk if the beneficiary’s costs exceed the capitated rate. About 125,000 dual-eligible military retirees reside in the catchment areas How the of the six sites—about one-fifth of dual eligibles living within 40 miles of Demonstration Works an MTF. About 30,000 will be allowed to enroll in the demonstration on a first-come, first-served basis. Demonstration participants will enroll in TRICARE Senior Prime, a new, DOD-run HMO exclusively for the demonstration areas and open to dual eligibles only. Senior Prime offers hospital, physician, and other Medicare-covered services. Senior Prime builds on TRICARE Prime, adding home health and other Medicare-required services. Under the demonstration, DOD will not charge enrollees a premium, at least for the first year. Services may, at Senior Prime’s option, be provided at an MTF or by a civilian network provider, but copayments differ by where the service is provided. For example, inpatient hospitalization will be free at the MTF but require a copayment for civilian providers. DOD anticipates that most services will be provided in MTFs. Like enrollees in private Medicare HMOs, Senior Prime enrollees are “locked out” of Medicare fee-for-service coverage. An enrollee who uses a civilian provider without a Senior Prime referral or authorization is Page 4 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 responsible for the full charge. Like commercial Medicare HMOs and other private, managed care plans available through Medicare+Choice, Senior Prime gets a capitated Medicare payment for each enrollee.5 In addition, Senior Prime must comply with all Medicare requirements for the protection of beneficiaries, provision of information, cost-sharing limitations, access, quality assurance, external review, and appeal and grievance procedures. Unlike a conventional Medicare+Choice plan, Senior Prime is established and operated by DOD; in addition to the standard benefits offered by a private Medicare+Choice plan, Senior Prime gives its members priority for treatment at MTFs over other dual eligibles.6 To be eligible for Senior Prime, a military retiree (or dependent or survivor) must: • be enrolled in both Medicare part A and part B (an estimated 90 percent of dual eligibles are enrolled in part B); • reside in one of the six geographic areas covered by the demonstration; • be a dual-eligible beneficiary who used an MTF before January 1, 1998, or became dually eligible (turned 65) after December 31, 1997; and • agree to use Medicare-covered and MTF services only through Senior Prime. The six sites for the demonstration differ considerably in their numbers of retired Medicare-eligible beneficiaries and in what DOD terms “enrollment capacity”—in effect, each site’s planned enrollment (see table 1). The sites also differ in several other ways, such as region, branch of service responsible for the MTF, size, and amount of managed care penetration in the local market. 5 The BBA provided that the demonstration rates be 95 percent of Medicare+Choice rates, adjusted to exclude payments for direct and indirect medical education and disproportionate share hospitals. Furthermore, the BBA also provided that a share of DOD’s capital costs be excluded from the rate, and that HHS and DOD must decide what that percentage share is. They have set the capital cost exclusion at 67 percent. 6 The subvention demonstration has a second component—Medicare Partners. Under Medicare Partners, a demonstration MTF can contract with Medicare+Choice plans to provide dual-eligibles enrolled in these plans with selected services at the MTF. It appears that generally MTFs will “sell” only specific services, such as the services of certain specialties, for which they have excess capacity. DOD agreed not to implement Medicare Partners until at least 90 days after the beginning of Senior Prime enrollment. It appears that it may be a year before Medicare Partners is activated at any site. Page 5 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 Table 1: Demonstration Sites for Medicare Subvention Start of service Current enrollment Site name delivery Dual eligiblesa Planned enrollment (as of 5/1/99)b Colorado Springs Evans Army Community Hospital, January 1, 1999 13,689 3,200 2,895 Ft. Carson; and 10th Medical Group, Air Force Academy, CO Dover 436th Medical Group, Dover Air Force January 1, 1999 3,905 1,500 678 Base, Dover, DE Keesler Keesler Medical Center, Biloxi, MS December 1, 1998 7,361 3,100 2,687 Madigan Madigan Army Medical Center, September 1, 1998 21,709 3,300 3,634 Ft. Lewis, Takoma, WA San Antonio San Antonio Wilford Hall Medical Center, Lackland October 1, 1998 34,148 10,000 10,413 Air Force Base; and Brooke Army Medical Center, Ft. Sam Houston, TX Texoma Reynolds Army December 1, 1998 7,067 2,700 1,844 Community Hospital, Ft. Sill, Lawton, OK; and Sheppard Air Force Base Hospital, Wichita Falls, TX San Diego Naval Medical Center San Diego, November 1, 1998 35,619 4,000 2,897 San Diego, CA Total 123,498 27,800 25,048 Note: A site may include more than one MTF and more than one geographic area. a Data are from the Defense Medical Information System for fiscal year 1998, third quarter. b Current enrollment can be more than planned because of “age-ins,” which are enrollees who reached age 65 after December 31, 1997. Source: DOD. The amount that Medicare will pay DOD for subvention enrollees depends not only on Medicare’s capitated rate for Senior Prime but also on DOD’s historical, or baseline, health care costs, termed LOE. The BBA required that DOD maintain its previous LOE in providing space-available care to dual-eligible retirees in the demonstration areas and that the Medicare payment reimburse DOD only for care above the LOE. As agreed by HCFA and Page 6 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 DOD, DOD cannot receive any Medicare payments unless current DOD expenses for the dual eligibles reach this baseline. Measurement of LOE is sensitive to data quality and reliability. If costs are omitted from LOE, DOD may be overpaid, but if LOE is inflated, Medicare will pay too little or perhaps nothing. Facility cost and workload data used to establish DOD’s LOE are drawn primarily from DOD’s Medical Expense Performance Reporting System (MEPRS). MEPRS data are used for many military health care services or management purposes such as resource allocation determinations, “make-versus-buy” decisions—such as whether to offer certain product lines or purchase them as needed, setting third-party billing rates, and cost comparisons of DOD’s health care delivery system with other alternatives. Thus, LOE accuracy and key military health care system functions rely in large measure on MEPRS and related data systems to provide accurate, timely, and complete cost and workload information. Portions of DOD’s LOE may be understated because of inaccuracies in its LOE Source Data source data, and as a result, Medicare overpayments may occur during the Inaccuracies May demonstration. DOD’s health care information systems are generally not Result in Medicare auditable and often cannot be reconciled with source data and documents. Military and civilian pay and prescription drugs exemplify areas of Overpayments possible inaccuracy. These problems stem from a long-standing lack of DOD and service oversight and incentives to ensure the data’s accuracy, timeliness, and completeness. In response to our preliminary findings, DOD recently developed a plan for improving MEPRS data and business practices both during and after the demonstration. The effects of these efforts on data quality and DOD’s ability to measure demonstration costs remain to be seen. In addition, DOD officials told us that they are committed to making any necessary changes to ensure that Medicare does not overpay DOD. Uncertainty About Data DOD has acknowledged concerns about MEPRS, its key system for estimating Quality Reduces costs for military health care. DOD officials described it as a “stepchild” Confidence in LOE system that has been underfunded and inconsistently used. As a result, DOD and the services have not effectively monitored MEPRS to ensure data Estimate quality. The MEPRS policy manual states that the Assistant Secretary of Defense (Health Affairs) is responsible for MEPRS direction and management; the DOD Comptroller is responsible for finance, budgeting, and accounting guidance for all health care resources; and the services are responsible for implementing MEPRS guidance and reporting uniform and Page 7 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 comparable data. But at the three sites we visited, we found that MTF staff did not fully audit MEPRS’ expense, workload, or manpower data for 1996 or later years. And recent DOD self-assessment surveys of MEPRS and other workload data quality showed wide variances among facilities. We and others have identified major concerns with MEPRS, including inconsistent data collection and reporting, service differences in how depreciation is recorded and what is counted as “readiness”7 (and thus not counted as patient care), and the completeness of the accounting for all relevant expenses. Responding to our questions and concerns, in 1998 DOD developed a MEPRS Management Improvement Plan. The plan focuses first on the subvention sites and turns to improving the entire system later. (See app. III.) The goal is for a working group composed of Army, Navy, and Air Force officials to develop and initiate standard business rules for recording, collecting, and reporting MEPRS data. The group is assessing the feasibility of incorporating into MEPRS other DOD appropriations—such as research, development, testing, and evaluation; military construction; military pay; and civilian pay—to capture all MTF revenues and expenses. Part of this plan calls for reconciliation of MEPRS data on finance, manpower, and workload with source documents and with data systems that provide information to MEPRS. However, DOD’s plan does not address some aspects of data quality.8 For example, we found evidence that basic data consistency checks had not been performed. Also, even when improved, MEPRS will continue to provide costs by cost center, functional area, and program, but not by individual patient or groups of patients. Consequently, an improved MEPRS may still not be ideally suited to identifying the costs of groups, such as the demonstration’s dual-eligible retirees. A DOD official told us the agency is planning to award a contract to determine how its systems compare with other health care cost systems; whether changes are needed; and, if so, the extent and feasibility of such changes. Problems in Estimating Our analysis showed that the demonstration may result in Medicare Major Cost Components overpayments. Two cases illustrate how data and related estimation Point to Potential Medicare problems may lead to a significant understatement of LOE. Overpayments 7 Readiness is the capacity to engage in military action. 8 The plan also does not address how MEPRS should be used for financial reporting, which requires the use of full cost accounting as defined by federal accounting standards. Page 8 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 The first concerns military and civilian pay. Military personnel account for more than half of total military health care expenses. However, all DOD activities, medical or otherwise, use service-specific composite pay rates—rather than actual pay—for estimating labor costs. This approach appears to understate actual pay at demonstration facilities. For example, DOD applies the same pay rate to a hospital administrator who is a lieutenant colonel and to an orthopedic surgeon of the same rank. This method would understate actual pay because actual salaries for physicians are generally higher than those for other personnel of the same rank. In particular, physicians receive larger and more frequent special pay allowances compared with nonphysicians. A study of Air Force MTFs by the Institute for Defense Analysis found that composite rates understated military physicians’ salaries but that these understatements were offset by an overestimate for nonphysicians.9 At large facilities, the understatement of physicians’ salaries would be expected to be greater, because these MTFs have more specialists. In the subvention demonstration, four of the MTFs, representing over 60 percent of the demonstration’s planned enrollment, are major medical centers—Madigan Army Medical Center (Wash.), Brooke Army Medical Center (Tex.), Wilford Hall Medical Center (Tex.), and Naval Medical Center San Diego (Calif.). In reviewing data from Wilford Hall Medical Center, we compared the national composite pay rate used in calculating LOE with another composite pay rate used locally and found a 6.8-percent difference. (We also found a 5.3-percent difference in civilian pay.) Our examination of the data did not provide grounds for choosing one rate over the other, but differences of this magnitude are cause for concern. DOD maintains that its composite pay rate approach is appropriate, because it reflects the way that appropriations for pay are distributed to all DOD facilities. Furthermore, DOD contends that collecting actual pay data would be costly. We believe, however, that while DOD’s composite pay rate approach may be accurate nationally and acceptable for other purposes, because all facility differences average out, it appears to understate actual pay at the subvention facilities by eliminating factors that make their personnel and compensation mix unique and above average. The treatment of physicians’ pay is the most pertinent, but other factors that may differentiate these particular facilities from the average facility, such as locality pay, are also omitted, except as they are reflected in national 9 Institute for Defense Analysis, Cost Analysis of the Military Medical Care System: Final Report, P-2990 (Washington, D.C.: Institute for Defense Analysis, Sept. 1994). Page 9 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 averages. DOD and HCFA have agreed to continue reviewing this issue and to make any needed changes. A second probable source of LOE understatement is the adjustment to exclude prescription drug expenses. Medicare generally does not cover outpatient prescription drugs, so the demonstration’s Memorandum of Agreement excludes prescription drug costs from the LOE for the six sites. However, DOD accounting systems often do not distinguish between pharmaceutical supplies used in clinic operations, such as chemotherapy drugs, and drugs that patients take home. This broad pharmaceutical category amounts to about $17 million in LOE (according to the DOD contractor responsible for estimating LOE). In removing all expenses in this category, not just those for outpatient drugs, DOD appears to be understating LOE. DOD has not offered a compelling reason for removing the entire amount from LOE. DOD officials have said that they will study this issue and make any necessary adjustments. Recent Changes in Data Improvements in DOD’s health care cost and information systems are likely Systems and Choice of to result in better measurement of current costs, but this may have a 1996 as Base Year Raise perverse effect on Medicare payments. If certain omissions or inaccuracies are left uncorrected in LOE but later corrected in current demonstration Concerns costs, the more accurately measured current costs will be tallied against the deficient baseline. This situation would make it easier for DOD to meet its LOE thresholds and tests, and thus to get Medicare payments (see app. I). Along with the health care cost data problems, we found that much of the documentation supporting the base year (1996) calculations is no longer available, hindering data verification. DOD and HCFA recently considered changing the LOE base year to 1998 because the data would be more readily available and auditable. However, DOD and HCFA have concluded thus far that the 1996 data may be adequate for the demonstration purpose and have not changed the demonstration baseline, although they continue to analyze the issue. HCFA Began Reviewing Although HCFA officials were involved in designing the demonstration, Baseline Data and including the Medicare payment provisions, annual reconciliation, and Methodology as a Result of beneficiary marketing processes, they had not reviewed DOD’s baseline data and methodology for compliance with the demonstration’s terms or Our Early Findings with Medicare reimbursement regulations until we disclosed our Page 10 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 preliminary findings. HCFA officials told us that the limited number of HCFA staff assigned to the demonstration have other responsibilities and thus have been unable to devote full attention to the project. In discussions with us, HCFA officials acknowledged that DOD’s LOE methodology and supporting data are more complex and problematic than they originally believed. They told us that they are assigning more staff to review the methodology and data; are committed to working with DOD to improve the LOE estimate; and are planning to award a contract to review all the issues we identified. Furthermore, DOD and HCFA officials told us they plan to continue meeting to clarify the Memorandum of Agreement’s details so that misunderstandings between the two agencies are minimized and the demonstration is implemented as efficiently as possible. For DOD, the real challenge of subvention is to establish and run a managed Data Weaknesses and care system that meets the requirements of Medicare and its beneficiaries. Payment Complexity To meet its responsibilities, DOD must manage the subvention Limit DOD in demonstration and track its progress toward reaching the LOE target. In addition, like other managed health care plans, DOD must manage costs Managing the and resources to maintain access to and quality of care. These are Demonstration and Its data-intensive tasks, and inadequate data systems will undermine a managed care plan’s ability to compete effectively. In addition, the Broader Health demonstration’s complex payment arrangements, and the fact that HCFA System and DOD have yet to specify a risk-adjustment method and how sites are to be paid, add uncertainty for DOD managers. Consequently, the inadequacies of DOD’s data systems limit its ability, at both the site and national levels, to manage the demonstration and deliver health care. Data Inaccuracies Hamper In taking responsibility for all Medicare-covered care of its Senior Prime DOD in Determining enrollees, DOD needs to know whether Medicare reimbursement covers DOD’s costs to deliver this care. DOD believes that its costs overall are less Whether Medicare than civilian costs, and an Institute for Defense Analysis study,10 which Reimbursement Covers compared peacetime military health care costs with civilian costs, partially DOD Costs and in supports that conclusion. However, the Institute for Defense Analysis Assessing Make/Buy study encountered considerable difficulties in using DOD data to determine Choices costs and made major adjustments to compensate for data limitations. The study found that DOD’s costs were about 6 percent less than the private sector’s. However, this estimate was based largely on data for a nonelderly population that would use fewer resources per person than retirees aged 10 Cost Analysis of the Military Medical Care System: Final Report (Sept. 1994). Page 11 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 65 and over, and it did not include the costs of providing skilled nursing facility and home health care. These two services account for about one-seventh of Medicare’s cost per beneficiary. DOD will need accurate, timely tracking of costs and utilization, particularly because 95 percent of the modified Medicare+Choice rates does not appear to leave DOD a large margin above cost. Like other managed care organizations, DOD continually makes decisions about whether to treat particular patients or send them to external network providers and whether to offer certain services or product lines or purchase them as needed. These decisions are usually made on the basis of incremental or marginal cost and may vary over time, depending on market conditions and other factors. It does not appear that MEPRS or other data systems currently give DOD adequate or accurate cost information on which to base these decisions. Some decisions are simple, of course—if a patient needs a kidney transplant and an MTF cannot provide it, the service must be purchased. But some represent choices between providing care in the MTF or in the community, such as whether to purchase some or all radiology services or provide them at the MTF. Inadequate cost data may lead MTF managers to select the more costly option. Payment Rules Create The payment arrangements of the demonstration complicate its operation Uncertainty for DOD (see app. I). DOD will not know until reconciliation takes place—roughly 6 Managers months after the year’s end—how much final payment it will receive. For care delivered in 1999, the annual reconciliation may not be completed until mid-2000. Adding to this uncertainty is that HCFA has not yet specified the method and criteria for adjusting Medicare payments for differences in enrollees’ health status. Furthermore, individual sites do not know, because DOD has not indicated, how money from Medicare will be distributed among the sites. In theory, DOD could give part of the final payment to sites according to their success in meeting monthly thresholds, or it could use the final payment to rescue less “successful” sites or to compensate sites that have sicker-than-average patients. Different payment scenarios will likely cause site managers to change their decisions about enrollment and capacity. For example, if DOD allows sites to spend all or part of interim (monthly) payments or allocates part of final payments to sites, site managers are likely to increase capacity and try to expand Senior Prime enrollment. This situation is less likely if DOD opts not to use a site’s performance to determine its share of final payments but Page 12 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 instead uses the payments to rescue less successful sites or for other purposes. In view of this uncertainty, some sites may pay for dual eligibles’ care exclusively from their site budgets, which draw on DOD’s appropriated funds. Because final payments from HCFA are determined in the year after care is delivered, Medicare funds cannot be relied on to pay for care. As a result, a site manager faced with expenses that threaten to exceed the site budget has three primary choices: reduce care for enrollees, reduce care for nonenrollees 65 and over and for younger military beneficiaries, or reduce enrollment through attrition. The extent to which such uncertainties will affect sites’ management of Senior Prime will be clearer after the sites have had more experience with subvention. Long-Standing Data DOD’s cost and workload reporting system weaknesses have effects that Problems Raise Broader reach beyond the Medicare subvention demonstration. These data are Concerns About System used throughout the military health care system by facility, service, and headquarters managers to make policy decisions, evaluate program Management effectiveness, and track expenditures against budgeted funds. But in recent years, we and others have identified data weaknesses that indicate limitations in DOD’s ability to, for example, • project accurate system costs for allocating resources; • establish accurate billing rates for third-party insurer collections that provide millions of dollars of revenue each year; and • conduct make-versus-buy analyses for improving the quality, accessibility, and cost-effectiveness of military health care—including weighing alternatives for providing beneficiaries’ care such as the Federal Employees Health Benefits Program. In addition to the MEPRS management improvement plan, DOD established a TRICARE data quality task force to address the broader system causes of the data problems that we and others continually have identified. DOD’s cost system problems are persistent and long-standing. In 1992, for example, DOD’s Office of Inspector General (OIG) reported that MEPRS did not track all costs associated with the delivery of peacetime health care, thereby understating the actual costs of operating and supporting MTFs. In addition, third-party billing rates did not reflect the total costs of the health care provided, resulting in understated billings. Also, health care cost information could not be easily retrieved and was not standardized, Page 13 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 and military composite rates did not reflect the actual labor costs of medical professionals.11 In 1995, the OIG reported on problems with the source systems underlying MEPRS. DOD’s general fund accounts, which are drawn upon to pay health care expenses, were not auditable because assets were not properly valued or reported in the accounts, contingent liabilities were not properly recognized or disclosed, disbursements and collections were not properly accounted for, and adequate accounting systems generally were not in place.12 Furthermore, in 1998, the OIG reported that data used to calculate the military retirement health benefits liability were neither current nor complete.13 Other studies by the OIG, contractors, and researchers during the period likewise identified and documented many of the same data inaccuracies and omissions in DOD’s health care information systems.14 Our review of the subvention baseline’s data reliability, moreover, has served to affirm that the data system problems identified over the years continue and thus affect all DOD health care operations that rely on these systems. DOD’s enrollment-based capitation program, for example, can be used to allocate resources to MTFs on the basis of their TRICARE enrollment levels and assign prices for an MTF’s services to be charged other MTFs when they refer patients to that facility. This “transfer pricing” portion of the enrollment-based capitation program relies heavily on cost data to calculate the payment to MTFs for their services. If underpaid, MTFs may experience funding shortfalls and be forced to restrict care. Thus, the enrollment-based capitation program’s implementation guidance stressed that MTF managers should make data quality a top priority if the program was to succeed. Moreover, DOD recently sought to reconcile MEPRS expenses with finance system obligations to correct data errors that would affect transfer prices. DOD found incomplete MEPRS data and mismatches in 11 DOD OIG, Peacetime Health Care Costs in the Military Health Services System, Report No. 92-PED-04 (Washington, D.C.: DOD OIG, Sept. 1992). 12 DOD OIG, Major Deficiencies Preventing Auditors from Rendering Audit Opinions on DOD General Fund Financial Statements, Report No. 95-301 (Washington, D.C.: DOD OIG, Aug. 1995). 13 DOD OIG, DOD Military Retirement Health Benefits Liability for FY 1997, Report No. 99-010 (Washington, D.C.: DOD OIG, Oct. 1998). 14 RAND National Defense Research Institute, Evaluation of the CHAMPUS Reform Initiative, Volumes 3 and 6, R-4244/3-HA and R-4244/6-HA (Santa Monica, Calif.: RAND, 1993 and 1994), and The Demand for Military Health Care: Supporting Research for a Comprehensive Study of the Military Health Care System, MR-407-1-OSD (Santa Monica, Calif.: Rand, 1995); DOD OIG, Review of Utilization Management in the Military Health Services System (Washington, D.C.: DOD OIG, June 1995), and Reporting Graduate Medical Education Costs, Report No. 97-147 (Washington, D.C.: DOD OIG, May 1997); and Institute for Defense Analysis and CNA Corporation, Evaluation of the TRICARE Program: FY 1998 Report to the Congress (Washington, D.C.: Institute for Defense Analysis and CNA Corporation, 1998). Page 14 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 facilities’ MEPRS and obligations data that would significantly understate prices established for those facilities. As a result, DOD urged MTF commanders to review and, to the extent possible, correct their MEPRS data. MEPRS data are also used to calculate MTFs’ third-party reimbursement rates. Such reimbursements include MTF collections from beneficiaries’ non-DOD health insurance policies. In fiscal year 1997, DOD collected almost $140 million in such reimbursements. These collections are projected to decline because many beneficiaries drop their third-party insurance after they enroll in TRICARE Prime. To ensure such collections are maximized, it is important that MTF billing rates accurately reflect the facilities’ costs. Weaknesses in DOD’s cost data can also impair the ability to evaluate alternate approaches to providing care to military beneficiaries. MTF commanders regularly confront make-versus-buy decisions and need reliable data to decide when to provide care at the MTF and when to seek private sector alternatives. Moreover, analyzing the cost-effectiveness and feasibility of new approaches—such as Medicare subvention, a mail-order pharmacy benefit for retirees, or Federal Employees Health Benefits Program coverage for senior retirees—also requires data on military facility care costs compared with these options. For example, the “733 Study,” DOD’s 1994 comprehensive study of military health care, drew heavily upon MEPRS data to compare DOD facility care costs with care provided under the Civilian Health and Medical Program of the Uniformed Services (CHAMPUS), the precursor to TRICARE.15 The study’s conclusion that DOD’s facility costs generally were lower has been challenged and today remains at issue. Therefore, DOD’s MEPRS cost and workload data should be as accurate as possible to support day-to-day system management and to provide the Congress with accurate assessments of system alternatives. As noted earlier, DOD established a high-level data quality task force to begin addressing what officials now see as an urgent need for data quality improvements. The task force’s mission statement reiterates that clinical workload data are used by DOD’s medical departments in their budgetary decisions, manpower justifications, program actions, and facility “rightsizing” initiatives. In addition, data-dependent managed care support contracts, enrollment-based capitation endeavors, and the Medicare subvention demonstration accentuate DOD’s reliance on accurate data. The 15 DOD, Office of Program Analysis and Evaluation, The Economics of Sizing the Military Medical Establishment, Executive Report of the Comprehensive Study of the Military Medical Care System (Washington, D.C.: DOD, Apr. 1994). Page 15 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 statement points out that data systems such as MEPRS, the Composite Health Care System, and the Ambulatory Data System, which support the MTFs in their daily activities, were developed independently and are not linked, leading to financial, workload, and data accuracy issues. Other contributing factors cited include • lack of consistent command emphasis to ensure that workload and other data reports are complete, timely, and accurate; • paucity of business rules, standardized training, and procedural guidelines for clerical and professional staff; • segmentation of functions and staffing as well as cultural and operational differences among the services and their facilities; and • conversion to a data-driven managed care environment involving new management methods that require accurate, relevant data. The mission statement establishes a December 2000 project completion date but notes that the project’s complexity and magnitude may require an extension. While this project is daunting, we agree it is critical that DOD begin to take actions needed to improve its data quality and that it fully commit itself to the project’s success. However, even if the target date is met, the project can have only limited impact on the subvention demonstration, which is scheduled to end at the same time. The Medicare subvention demonstration provides DOD and HCFA a valuable Conclusions opportunity to gauge the effects of treating Medicare-eligible beneficiaries in military facilities. However, the demonstration’s payment rules and method of estimating LOE demand accurate, timely, and complete data, and DOD’s ability to provide such information with its current systems is questionable. These data problems also call into question DOD’s ability to manage its overall health care system. In short, DOD lacks an information system that can produce credible cost data on its individual beneficiaries and beneficiary groups. Yet, even with good information systems, DOD and the demonstration sites face a considerable challenge in managing the demonstration. For example, the demonstration sites will not know how much they will be paid for a given year until well into the following year, and DOD has not yet made other decisions regarding sites’ interim and final payments. Nonetheless, the experience of the demonstration will provide valuable information for developing a permanent reimbursement system, if the demonstration is deemed to meet its cost, quality, and other goals. Page 16 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 Beyond the demonstration, DOD’s many other needs for reliable cost data warrant that it dedicate sufficient effort to improving the data’s accuracy and reliability. In that regard, DOD’s new management improvement plan and data quality task force are positive steps. We urge DOD’s continued high-level attention to these issues. We recommend that the Secretary of Defense direct the Assistant Recommendations to Secretary of Defense (Health Affairs), in collaboration with HCFA, to the Secretary of identify the baseline’s weaknesses and resulting errors in LOE and Defense determine a more reliable baseline. This effort should consider the merits of using a more recent base year for the demonstration and weigh alternatives to the current baseline method. Furthermore, to reduce funding uncertainties for site managers, the Assistant Secretary should state definitively how final Medicare payments will be allocated among the demonstration sites, and working with HCFA, explain the method and criteria for risk-adjusting sites’ Medicare payments. We also recommend that the Secretary of Defense direct the Assistant Secretary of Defense (Health Affairs) to improve cost and workload data quality. This is especially important because DOD also uses these data in managing its general health care operations. The effort should identify specific actions needed by the Assistant Secretary and the services to correct current cost and workload data collection and reporting problems. It should also ensure, by maintaining all source data and documents, that MEPRS can be audited. This effort may require actions by and coordination with other DOD Assistant Secretaries, and the Secretary should direct their participation. We recommend that the Administrator of the Health Care Financing Recommendation to Administration, in collaboration with DOD, identify the baseline’s the Administrator of weaknesses and, as appropriate, determine a more reliable baseline. HCFA HCFA efforts should include providing DOD specific guidance on baseline cost components and assessing baseline source data and methods for reliability and compliance with HCFA guidance and regulations. Also, working with DOD, the Administrator should promptly specify the method and criteria for risk-adjusting the Medicare payments. DOD and HCFA commented on a draft of this report. DOD found the report Agency Comments valuable in raising issues deserving its immediate attention; agreed with and Our Evaluation Page 17 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 each recommendation for its action; and stated that it would continue working with HCFA to improve the measurement of LOE and improve its data systems. HCFA stated that after we made our preliminary findings known last year, it began working closely with DOD on the LOE data accuracy issues. HCFA stated that while both parties have agreed thus far to keep the 1996 baseline, it was awarding a contract to review the threshold’s weaknesses and identify needed improvements. Also, as we recommended, HCFA stated that both parties now agree on a payment reconciliation approach that will be made final shortly. Both parties also suggested technical changes to the report, which we incorporated where appropriate. DOD and HCFA comments appear in their entirety in appendixes V and VI, respectively. We are sending copies of this report to the Honorable William S. Cohen, Secretary of Defense, and the Honorable Nancy-Ann Min DeParle, Administrator of HCFA, and will make copies available to others upon request. Please contact me at (202) 512-7111 or Dan Brier, Assistant Director, at (202) 512-6803 if you or your staff have any questions about this report. Other GAO staff who contributed to this report are Catherine O’Hara, Evaluator-in-Charge; Linda Radey; Jonathan Ratner; Phyllis Thorburn; and Sibyl Tilson. Stephen P. Backhus Director, Veterans’ Affairs and Military Health Care Issues Page 18 GAO/HEHS-99-39 DOD/Medicare Subvention B-278140 List of Addressees The Honorable John W. Warner Chairman The Honorable Carl Levin Ranking Minority Member Committee on Armed Services United States Senate The Honorable William V. Roth, Jr. Chairman The Honorable Daniel Patrick Moynihan Ranking Minority Member Committee on Finance United States Senate The Honorable Floyd D. Spence Chairman The Honorable Ike Skelton Ranking Minority Member Committee on Armed Services House of Representatives The Honorable Tom Bliley Chairman The Honorable John D. Dingell Ranking Minority Member Committee on Commerce House of Representatives The Honorable Bill Archer Chairman The Honorable Charles B. Rangel Ranking Minority Member Committee on Ways and Means House of Representatives Page 19 GAO/HEHS-99-39 DOD/Medicare Subvention Contents Letter 1 Appendix I 22 Process for Determining Medicare Payments Appendix II 24 Scope and Methodology Appendix III 25 DOD’s MEPRS Management Improvement Plan Appendix IV 28 Establishment of Health Care Data Quality Team Appendix V 33 Comments From the Department of Defense Appendix VI 35 Comments From the Health Care Financing Administration Table Table 1: Demonstration Sites for Medicare Subvention 6 Page 20 GAO/HEHS-99-39 DOD/Medicare Subvention Contents Abbreviations BBA Balanced Budget Act of 1997 CHAMPUS Civilian Health and Medical Program of the Uniformed Services DOD Department of Defense HCFA Health Care Financing Administration HMO health maintenance organization LOE level of effort MEPRS Medical Expense Performance Reporting System MTF military treatment facility OIG Office of Inspector General Page 21 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix I Process for Determining Medicare Payments The process for determining whether and how much Medicare pays to DOD under the demonstration program begins with the estimation of DOD’s level of effort (LOE), or baseline costs. To derive LOE, DOD estimated its actual expenses in providing care to retirees during a base, or reference, year. Using 1996 as the base year, DOD currently estimates LOE for the six sites at $172 million. As agreed by HCFA and DOD, Medicare payment does not start until current expenses reach this baseline. Thus, to the extent DOD’s baseline expenses may be over- or understated, Medicare either will under- or overpay. And, if expenses are captured during the demonstration that were not included in the baseline, the baseline costs will be reached more easily, which will erroneously trigger payments. Medicare payments to DOD involve both interim reimbursement, which is monthly, and an annual reconciliation to determine final payment. DOD will receive interim payments from Medicare that are based on monthly site LOE thresholds. Only when a site’s enrollment in Senior Prime meets a specified threshold, which is a percentage of the site’s LOE, will interim payments be triggered.16 The site is not required to meet the annual threshold—a percentage of the annual LOE—before it is entitled to interim payments. At the end of the year, two tests are applied to determine how much, if any, of the interim payments DOD can retain. First, expenses for all dual eligibles (enrollees and nonenrollees) at all sites must meet or exceed LOE ($172 million). Second, expenses for enrollees (as proxied by capitated payments for them) must reach or exceed fixed thresholds—30 percent of LOE in the first year, 40 percent in the next year, and 50 percent in the third. If DOD passes these two tests, two additional steps determine the final payment. First, Medicare’s capitated rate, which is a modified version of the Medicare+Choice rate,17 is based on the average cost for Medicare enrollees by county. HCFA will “risk adjust” this rate for Senior Prime enrollees, raising the rate if the enrollees were sicker than average and decreasing the rate if they were healthier. Enrollees’ rates are not changed if their health was average. The Memorandum of Agreement signed by HCFA and DOD for the demonstration states that risk adjustment will take 16 Monthly interim payments are the capitated payments for all enrollees at a site minus the site’s monthly threshold. The threshold is the site’s monthly LOE multiplied by a stated percentage (30 percent in the first 10 months of the demonstration, 40 percent in the next 9, and 50 percent in the final 9 months). 17 Medicare’s rate for Senior Prime enrollees is 95 percent of the Medicare+Choice rate, with certain exclusions as specified in the Balanced Budget Act and the Memorandum of Agreement. Page 22 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix I Process for Determining Medicare Payments place only if the evidence of differences in health status is “compelling”; neither the method of adjustment nor the criteria for distinguishing compelling evidence from less convincing evidence are given. The second step requires, for each site, an offset to interim payments to account for any months in which enrollment fell short of the site’s threshold. Finally, expenses for space-available care are added to the capitated payments and baseline LOE is subtracted—the result is the final payment to DOD. The Balanced Budget Act caps payments to DOD at $50 million in the demonstration’s first year, $60 million in the second year, and $65 million in the third year. Page 23 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix II Scope and Methodology In conducting our evaluation, we reviewed the method for measuring DOD’s LOE and ongoing DOD health care costs for Medicare-eligible military retirees in the demonstration; we also reviewed key DOD information and accounting systems and the data drawn from these systems. We visited three MTFs—Brooke Army Medical Center and Wilford Hall Air Force Medical Center in San Antonio, Texas, and Naval Medical Center San Diego, California. These three centers are expected to account for more than half the workload in the six-site demonstration. While at these sites, we interviewed command, finance, and accounting staff, and reviewed cost and workload data. We also reviewed LOE cost calculations and interviewed DOD and HCFA officials responsible for the subvention demonstration. In addition, we visited Madigan Army Medical Center, Fort Lewis, Washington, shortly after it began delivering care under the demonstration. We also conducted an in-depth review of data system documentation and Office of Inspector General and other studies related to the quality of DOD data systems. The data system documentation we examined, our discussions with cognizant officials, and our review of other studies confirmed that the data systems used in estimating LOE and measuring ongoing DOD health care costs are also used to support DOD health budgetary and program decisions, manpower justifications, facility “rightsizing” initiatives, and managed care support contract payments. On this basis, we believe that our findings about DOD’s data are applicable systemwide. Page 24 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix III DOD’s MEPRS Management Improvement Plan Page 25 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix III DOD’s MEPRS Management Improvement Plan Page 26 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix III DOD’s MEPRS Management Improvement Plan Page 27 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix IV Establishment of Health Care Data Quality Team Page 28 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix IV Establishment of Health Care Data Quality Team Page 29 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix IV Establishment of Health Care Data Quality Team Page 30 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix IV Establishment of Health Care Data Quality Team Page 31 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix IV Establishment of Health Care Data Quality Team Page 32 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix V Comments From the Department of Defense Page 33 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix V Comments From the Department of Defense Page 34 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix VI Comments From the Health Care Financing Administration Page 35 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix VI Comments From the Health Care Financing Administration Page 36 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix VI Comments From the Health Care Financing Administration Now on p. 4. Now on p. 5. Now on p. 9. Paragraph deleted. Paragraph deleted. Now on p. 10. Now on p. 10. Page 37 GAO/HEHS-99-39 DOD/Medicare Subvention Appendix VI Comments From the Health Care Financing Administration (101607) Page 38 GAO/HEHS-99-39 DOD/Medicare Subvention 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 37050 Washington, DC 20013 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 (202) 512-6061, or TDD (202) 512-2537. 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
Medicare Subvention Demonstration: DOD Data Limitations May Require Adjustments and Raise Broader Concerns
Published by the Government Accountability Office on 1999-05-28.
Below is a raw (and likely hideous) rendition of the original report. (PDF)