|
United States General Accounting Office Testimony Before the Subcommittee on Oversight and Investigations, Committee on Veterans’ Affairs, House of Representatives VETERANS BENEFITS ADMINISTRATION Problems and Challenges Facing Disability Claims Processing Statement of Cynthia A. Bascetta, Associate Director Veterans' Affairs and Military Health Care Issues Health, Education, and Human Services Division Veterans Benefits Administration: Problems and Challenges Facing Disability Claims Processing
Mr. Chairman and Members of the Subcommittee: We are pleased to be here today to provide an overview of claims processing in the disability compensation program through which the Department of Veterans Affairs (VA) provides cash benefits to more than 2.5 million veterans, their dependents, and survivors. The compensation program pays monthly benefits—based on degree of disability—to veterans who have service-connected disabilities (injuries or diseases incurred or aggravated while on active military duty). Administered by the Veterans Benefits Administration (VBA), the compensation program is VBA’s largest program, accounting for about 72 percent of fiscal year 1999 cash outlays (about $18 billion out of $25 billion). For years, the compensation program has been the subject of concern and attention within VA and by the Congress and veterans’ service organizations. The concerns have focused on backlogs of claims, long waits for disability decisions, and the poor quality of these decisions, all of which have negatively affected the quality of service provided to veterans. We have issued a number of reports on VBA’s claims-processing operations, and the Congress has sponsored studies of the disability compensation program, including studies by the Veterans’ Claims Adjudication Commission and the National Academy of Public Administration (NAPA). Today, drawing on this body of work, I will focus on four key areas related to compensation claims processing: (1) long-standing performance problems, (2) claims-processing complexities, (3) challenges to improving performance, and (4) VBA’s initiatives to improve performance. In summary, VBA’s problems with large backlogs and long waits for decisions have not yet improved, despite years of studying these problems. Moreover, VBA's new quality measurement system shows that nearly one-third of decisions are incorrect or have technical or procedural errors. Many performance problems stem from the process's complexity, which is growing as the number of service-connected disabilities per veteran increases and judicial review requires more procedures and documentation. Although VBA has initiated a number of efforts to streamline its claims-processing performance, it is unclear how much improvement will be gained. Also, VBA may need to collect and analyze additional case-specific data to better understand its claims-processing problems and better target its corrective actions. Furthermore, because some issues affecting VBA’s performance are a function of program design, more fundamental changes may have to be considered to realize significant improvements. Veterans may submit claims to any one of VBA’s 57 regional offices. To develop a veteran’s claim, the regional office obtains the veteran’s existing medical and military service records and, if necessary, arranges for the veteran to be examined by physicians in the Veterans Health Administration (VHA). The regional office evaluates the veteran’s service-connected impairments and assigns a rating for the degree to which the veteran is disabled, ranging from zero to 100 percent (expressed in 10-percent increments). For veterans with multiple disabilities, the regional office combines the ratings for each disability into a single, composite rating. If a veteran disagrees with the regional office's decision, he or she can ask for a regional office hearing or submit a "notice of disagreement" and file an appeal asking VA’s Board of Veterans’ Appeals to review the decision. The Board makes the final decision on such appeals and can grant benefits, deny benefits, or remand (return) the case to the regional office for further development and reconsideration. After reconsidering a remanded decision, the regional office either grants the claim or returns it to the Board for a final VA decision. If the veteran disagrees with the Board’s decision, he or she may appeal to the U.S. Court of Appeals for Veterans Claims. If either the veteran or VA disagrees with this court’s decision, they may appeal to the Court of Appeals for the Federal Circuit. For a number of years, VBA's regional offices have experienced problems processing compensation claims. These have included large backlogs of pending claims, lengthy processing times for initial claims, high error rates in claims processing, and questions about the consistency of regional office decisions. As acknowledged by VBA, backlogs of claims have resulted in veterans having to endure long waits to receive decisions on their initial claims and on their appeals. As shown in figure 1, at the end of fiscal year 1999, VBA had about 69,000 pending initial compensation claims, of which over 23,000 (34 percent) had been pending for more than 6 months. You can see that in all categories the number of claims pending has been growing since 1996. Figure 1: Initial Compensation Claims Pending at Year-End, FY 1995-99
Source: VBA data. The average time for processing initial compensation claims peaked at 213 days in fiscal year 1994, as shown in figure 2. Thereafter, timeliness seems to improve through fiscal year 1997, as average processing time declined to 133 days. However, according to VA, apparent improvements were based on timeliness data that substantially understated the actual time required to process claims. This was revealed by a VA Inspector General audit, which found that timeliness data reported by regional offices had been in error by as much as 34 percent. After VBA took action to correct the data reporting problems, the average processing time again climbed, reaching 205 days in fiscal year 1999. This places VBA far from reaching its strategic goal of 74 days average processing time for claims that require disability ratings. Figure 2: Average Processing Time (in Days) for Initial Compensation Claims, FY 1990-99
Source: VBA data. When veterans appeal decisions made by regional offices, the average time spent to resolve the appeals is even longer than the time that the regional offices spent making the initial decisions. For appeals resolved during fiscal year 1999, the average time required was over 2 years (745 days) from the date the veteran submitted a notice of disagreement with the regional office’s decision. In addition to problems with timeliness of decisions, VBA acknowledges that the accuracy of regional office decisions needs to be improved. VBA historically had reported that regional offices processed claims accurately over 95 percent of the time; however, concerns about accuracy arose in the 1990s when dramatic increases occurred in the percentage of appealed cases remanded to regional offices by the Board of Veterans’ Appeals. As a result, VBA implemented a new accuracy measurement system in fiscal year 1999 under which the error rate includes not only incorrect decisions on whether to grant or deny claims but also procedural and technical errors such as failure to include all required documentation in the case file or to properly notify veterans of decisions. Using the new method, VBA calculated an accuracy rate of 68 percent (32-percent error rate) for initial decisions requiring disability ratings. For fiscal year 2000, VBA has set an accuracy goal of 81 percent; its long-term strategic goal is 96 percent accuracy. Another problem is the perception of inconsistency in decisions made by different regional offices. In 1997, NAPA identified several factors that could lead to inconsistency in VBA's decisions: (1) achieving consistency across 57 decentralized offices is inherently difficult, (2) regional office staff must deal with a variety of medical issues that often require them to make subjective judgments, (3) VBA’s regulations were unclear and subject to varying interpretations, and (4) VBA lacked a comprehensive training strategy that identified training needs and used standardized training to meet these needs. NAPA stated that VBA needed to identify the degree of subjectivity expected for various medical issues, set consistency standards, and measure the level of consistency as part of the quality review process or through testing of control cases in several regional offices. Regional offices perform six basic functions in processing initial claims for service-connected disability compensation. Although VBA has made some changes in the process and plans to make additional changes, regional offices will still need to perform the six basic functions:
As we reported in 1994, many in VA blamed part of the claims-processing delays on the traditional, assembly line processing approach used in regional offices. Under the traditional approach, each claim passed sequentially through several individuals who separately performed the six processing functions mentioned. VBA has started moving toward a team-based, case management approach under which a regional customer service team is collectively responsible for processing each claim from beginning to end, thereby avoiding multiple handoffs of the claim to individuals who separately perform each task. The regional offices are in various stages of implementing this new approach. In addition, the regional offices have implemented two systems to assist them with their work. One tracks the location of claims folders, while the other system prevents the entry of duplicate requests for service verification and service medical records. Also, for claimants discharged from military service after May 1, 1994, the Department of Defense now automatically transfers their service medical records to VA, alleviating the need to request these records. The changes made to date, however, have done little to streamline the overall process. Currently, the process contains as many as 66 decision points and 39 queues (or waiting points) (see the app. for a depiction of the initial compensation claims process). Of the 39 queues, 28 are points at which claims wait for attention from regional office staff, and 11 are points at which regional office staff wait for information from external sources not under their control. For example, NAPA reported in 1997 that it was not unusual for regional offices to take as long as 80 days to request and obtain information such as (1) military service dates; (2) service medical records; (3) verification of receipt and amounts of military severance pay, separation pay, and/or retired pay; (4) medical records from private physicians, hospitals, and VA medical centers; and (5) other evidence in the custody of military authorities or other government agencies. Even after obtaining this information, regional staff often find they need additional medical evidence to determine a veteran’s precise current medical status. In such cases, the staff must schedule the veteran for an examination by a VHA or contract physician. If regional staff find that the physician’s initial examination is not adequate, they must request a follow-up examination. Another factor that can increase complexity and contribute to claims-processing delays is that veterans have the right, by law, to submit additional evidence at any point during VA’s initial claims process, including during appeals on these claims to the Board of Veterans’ Appeals. The submission of such evidence can result in delays because claims processors must further develop the claim and reevaluate the veteran’s degree of disability. In addition to the claims-processing system itself, VBA faces challenges to its efforts to improve timeliness and accuracy in claims processing. These include (1) claims characteristics that increase workloads, such as the number of disabilities claimed by veterans; (2) decisions by the U.S. Court of Appeals for Veterans Claims that expand claims-processing requirements; and (3) a significant number of retirements by experienced staff that will require VBA to train many new employees. Veterans seeking compensation benefits often claim multiple disabilities. For example, in a sample of about 69,000 veterans whose initial claims were rated during fiscal year 1998, VBA found that the veterans claimed a total of about 316,000 disabilities, or an average of about 4.6 disabilities per veteran; the largest number of disabilities claimed by an individual veteran was 56. To process these claims, regional office staff had to make about 316,000 separate decisions that required development of evidence; determination of whether the disability was service-connected; and, if the disability was found to be service-connected, evaluation of the degree of disability. The number of disabilities determined to be service-connected has also been increasing. Of all the veterans who began receiving compensation benefits during fiscal year 1998, the average veteran had 2.72 service-connected disabilities. Compared with 1989, this represents an increase of about 30 percent in the number of service-connected disabilities per veteran. The increase in the average number of service-connected disabilities per veteran may be due to several factors. For example, NAPA commented on the possible effects of VA’s cooperative effort with the Department of Defense to perform medical examinations of veterans before their discharge from the service and to begin the claims process closer to the time of discharge. NAPA raised the possibility that these efforts potentially could result in the identification of a greater number of disabilities. The increase in disabilities per veteran also may be attributable in part to the recognition of new disabilities that are more difficult to evaluate. For example, the Agent Orange Act of 1991 presumed that anyone who served in Vietnam had been exposed to Agent Orange and extended compensation for certain diseases presumed to result from exposure. In another instance, the Veterans’ Benefits Improvement Act of 1994 identified Gulf War Syndrome as a compensable disability, which was the first time the Congress authorized VA to compensate veterans for "undiagnosed illnesses" for which only symptoms can be discerned. VBA data show that Gulf War veterans have more service-connected disabilities than any other group of veterans since World War II. Another factor that drives regional office workloads is "repeat" (or subsequent) claims filed by veterans after their initial claims are decided. According to VBA, repeat claims include requests for reevaluation of disabilities previously claimed or the evaluation of new disabilities not claimed previously. In fiscal year 1998, veterans filing repeat claims outnumbered veterans filing initial claims by about three to one. Additionally, as mentioned, the number of service-connected disabilities per veteran has been increasing. This increases the potential for repeat claims because each additional disability represents the potential for a request for reevaluation. Until the passage of the Veterans’ Judicial Review Act in 1988, decisions by VA’s Board of Veterans’ Appeals were not subject to judicial review. The act, however, established the U.S. Court of Veterans Appeals (now known as the U.S. Court of Appeals for Veterans Claims) and gave veterans the right to appeal the Board’s decisions to the Court. As the Board found its own decisions being remanded by the Court, the Board in turn began remanding many more cases to the regional offices for rework. (As mentioned, not every remand indicates that the regional office made an error.) Before the Court was established, the Board annually had remanded less than 25 percent of the cases it reviewed; however, after the creation of the Court, the proportion of cases remanded by the Board reached as much as 50 percent. Recently, the remand rate has declined—for the first 4 months of fiscal year 2000, the remand rate was about 29 percent, according to VBA and Board officials. Perhaps more importantly, the Court’s decisions also contributed to substantial increases in the time required to process claims. According to the Veterans’ Claims Adjudication Commission, VA historically has lacked clear and definitive administrative procedures, but prior to creation of the Court, VA’s vague rules had not been a problem because the rules were subject only to VA’s interpretation. The Court’s interpretation, however, of statutory and regulatory provisions generally has been more expansive than VA’s and has imposed greater procedural and documentation requirements on VA. For example, before the Court’s creation, regional office staff generally wrote one brief statement for each claim that summarized their overall evaluation and rating of all disability issues. Now, regional staff must separately describe the evidence and the decision rationale for each disability issue. The Adjudication Commission’s 1996 report stated that the number of work hours required to process the average case had doubled since the creation of the Court. Consistent with this finding, VBA data show that the number of decisions produced per rating specialist in fiscal year 1999 (797 decisions) was less than half the number produced 10 years earlier in fiscal year 1989 (1,716 decisions). According to VBA, it takes 2 to 3 years of experience for claims decisionmakers to achieve a fully productive level of expertise. Currently, about half of such VBA staff have 3 years or less of decision-making experience. The proportion of less experienced decisionmakers is likely to increase in the near future because of the expected retirement of over 1,100 experienced decisionmakers in the next 5 years. In the current fiscal year, VBA will add 440 new staff to the compensation and pension programs. In fiscal year 2001, VBA plans to redirect 183 existing staff positions to compensation and pension claims processing and hire 243 new staff. This highlights the need for an effective claims-processing training program. VBA has acknowledged that its training program has not adequately prepared its workforce to produce accurate disability decisions, and VBA has recognized the need for an effective, centralized, and comprehensive training program. VBA has acknowledged the need to improve the timeliness and accuracy of claims processing. Accordingly, VBA has an ongoing effort to reengineer the initial disability claims process as well as other initiatives aimed at improving performance. At this point, however, VBA’s initiatives are in various stages of testing and implementation, and it is not clear whether or to what extent these initiatives will improve timeliness or accuracy. Also, in some cases, VBA may need additional data to identify the underlying causes of its claims-processing problems. For example, as we reported in March 1999, VBA could further improve its claims-processing accuracy measurement system by collecting more specific data that would help identify error-prone cases and target corrective actions. VBA’s initiatives for improving claims processing encompass efforts such as implementing a case management approach for processing claims; working with the Department of Defense to administer physical examinations before servicemembers are discharged from military service; using electronic networks to obtain existing military service and medical records; improving the guidance and training for VHA physicians; developing computer-based training modules for regional office staff; and instituting a "balanced scorecard" that measures program performance on the basis of claims-processing accuracy and timeliness as well as unit cost, customer satisfaction, and employee development. In addition, during fiscal years 1986 through 1999, VBA spent at least $380 million to modernize its information technology systems to support its operations. Of the $380 million, at least $28 million was spent on initiatives specifically intended to improve compensation claims processing, from the establishment of claims through benefit payment and accounting. These initiatives are at various stages of completion. For example, in 1996 VBA implemented an initiative to track the location of veterans’ claims folders. Since then, VBA has been developing a system to replace the compensation and pension payment system. Also, in February 1999 VBA began testing the use of a case management approach to claims processing at six demonstration sites. As part of this test, VBA is using two automated tools: (1) the Claims Processing System applies rule-based technology to identify necessary evidence when a claim is initially received and produces reader-friendly letters requesting evidence and (2) the Claims Automated Processing System collects and stores information about pending claims. In August 1999 VBA completed a 6-month assessment of the demonstration project and concluded that neither system had any discernible effect on performance measures such as pending workload, timeliness, and productivity. VBA found that the Claims Processing System was labor intensive and had system access problems. It also found that the Claims Automated Processing System could not produce some management reports; this problem, according to VBA, has been fixed. According to a recent status report on its efforts to reengineer claims processing, VBA plans to continue using the Claims Automated Processing System to assist employees in providing case management services, but VBA discontinued the mandatory use of the rule-based Claims Processing System at the demonstration sites. Despite VBA’s efforts to improve its performance, its timeliness problems in claims processing continue and its accuracy in claims processing has far to go to reach VBA’s strategic goal for accuracy. At present, it is unclear how much improvement will be gained through VBA’s initiatives. Also, while VBA has improved its data collection efforts, it may still need to collect and analyze additional data, such as specific information on error-prone cases, to further understand its claims processing problems and better target corrective actions. Furthermore, as we mentioned in last year’s testimony before the Subcommittee on Benefits, some issues affecting VBA’s performance are not in its direct control and are a function of the design of the program. As a result, it may be that only incremental gains can be made without changes in the current design of the program. Mr. Chairman, this concludes my prepared remarks. I would be pleased to respond to any questions you or Members of the Subcommittee may have. For future contacts regarding this testimony, please call Cynthia A. Bascetta at (202) 512-7101. Others who made key contributions to this testimony are Irene Chu, Tonia Johnson, Helen Lew, Steve Morris, Barbara Oliver, Martin Scire, Ira Spears, Henry Sutanto, and Paul Wright. Figures 3a through 3i depict the initial compensation claims process. A list of abbreviations and forms referred to in the figures is included after figure 3i. Figure 3a: Process Legend
Figure 3b: Receive a Claim
Figure 3c: Establish a Claim
Figure 3d: Develop a Claim (Part 1)
Figure 3e: Develop a Claim (Part 2)
Figure 3f: Rate a Claim
Figure 3g: Determine Payment Amount
Figure 3h: Authorize a Claim
Figure 3i: File Banks
Orders by Internet For information on how to access GAO reports on the Internet, send an e-mail message with "info" in the body to: Info@www.gao.gov or visit GAO’s World Wide Web home page at: http://www.gao.gov Contact one: Web site: http://www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov 1-800-424-5454 (automated answering system)
Back to Witness List |
