RONALD R. AUMENT
DEPUTY UNDER SECRETARY FOR BENEFITS
DEPARTMENT OF VETERANS AFFAIRS
SUBCOMMITTEE ON DISABILITY ASSISTANCE AND MEMORIAL
HOUSE COMMITTEE ON VETERANS' AFFAIRS
October 20, 2005
Mr. Chairman and members of the Subcommittee: Thank you for the opportunity to review with you variances in disability compensation claims decisions made by VA regional offices, factors affecting our claims decisions, and recommendations for standardizing the adjudicative process. I am pleased to be accompanied by Ms. Renée Szybala, Director of VA’s Compensation and Pension Service.
I will also today discuss the November 2004 report by the Government Accountability Office (GAO) and the May 2005 report of the Office of the Inspector General (IG). Finally, as requested, I will provide our views on the United States Court of Appeals for the Federal Circuit holding in Allen v. Principi.
In October 2001, the VA Claims Processing Task Force, established by then Secretary of Veterans Affairs Anthony J. Principi, delivered its report containing 34 recommendations to improve compensation and pension claims processing.
The Task Force, which was chaired by Admiral Cooper, found that the most significant issue to be addressed – that would bring the most improvement to the decision-making process – was the need for greater accountability and consistency in our benefits delivery operations. Over the last 3½ years during the Admiral’s tenure as Under Secretary for Benefits, the Veterans Benefits Administration (VBA) has worked hard to address this need.
Through the implementation of the Task Force recommendations, VBA has achieved major improvements in the delivery of benefits including the quality of our benefits decisions – and we have laid the basic groundwork that will continue to bring more consistency in our decisions as well.
First, we have made all regional offices consistent in organizational structure and work process. Work processes were reengineered and specialized teams established to reduce the number of tasks performed by decision-makers, establish consistent work processes, and incorporate a triage approach to incoming claims.
Second, specialized processing initiatives have been implemented to consolidate adjudication of certain types of claims to provide better and more consistent decisions. Three Pension Maintenance Centers were established to consolidate the complex and labor-intensive work involved in ensuring the continued eligibility and appropriateness of benefit amounts for pension recipients. We are exploring the centralization of all pension adjudications to these Centers.
A Tiger Team was established to adjudicate the claims of veterans age 70 and older, and VBA established an Appeals Management Center to consolidate expertise in processing remands from the Board of Veterans' Appeals. In a similar manner, a centralized Casualty Assistance Unit was established to process all in-service death claims. Most recently, VBA has consolidated the rating aspects of our Benefits Delivery at Discharge initiatives, which will bring greater consistency of decisions for newly-separated veterans.
To further our drive toward consistency, we have established an aggressive and comprehensive program of quality assurance and oversight to assess compliance with VBA claims processing policy and procedures and assure consistent application. Included in this effort are oversight reviews, regularly performed by Headquarters staff. The Area Directors perform oversight visits as well. Training is provided, when appropriate, to address gaps. Accuracy reviews, statistically valid for each regional office, are provided by Headquarters. Centralized oversight, focused on quality and accuracy, will necessarily increase consistency as well.
We are working with the Veterans Health Administration through our joint VBA/VHA Compensation and Pension Examination Project (CPEP) Office to improve both the quality of examination requests from VBA regional offices and the examinations conducted by VA examiners. The examination is central to the ultimate evaluation decision. That effort is focused on two critical elements. First, our goal is to insure the examination request issued by the regional office is clear and comprehensive. Our second goal, through the development of templates, is to assure that the examination produced is complete, addresses all relevant elements such as the DeLuca criteria, and meets the needs of the rating specialist.
Training is central to every quality organization. VBA has deployed new training tools and centralized training programs that support greater consistency in decisions. New hires receive comprehensive training and a consistent foundation in claims processing principles through a national centralized training program called “Challenge.” After the initial centralized training, employees follow a national standardized training curriculum (full lesson plans, handouts, student guides, instructor guides, and slides for classroom instruction) available to all regional offices. Additionally, standardized computer-based tools have been developed for training decision-makers (53 modules completed and an additional 38 in development). Additionally, a policy mandating job-specific training hours for each employee will be implemented. Finally, training letters and satellite broadcasts on the proper approach to rating complex issues have been provided to the field stations.
Consistent utilization of information technology (IT) applications, key to developing usable data to monitor our progress, and consistent work-management systems are now required. Organizational and individual accountability has been established at all levels through consistent measures of performance and implementation of national performance standards.
VBA was very attuned to the consistency issue prior to receiving GAO’s November 2004 review of consistency in decision-making by VA regional offices. This was not the first time that GAO looked at this issue. Previous GAO studies in 2000 and 2002 raised concerns about the element of potential variability due to the nature of VA claims adjudication, which requires the application of judgment by the decision-maker. Judgment is recognized as crucial when assessing the credibility of different sources of evidence, weighing the comparative value of evidence developed, and assessing disabilities where the evaluative criteria are not entirely objective. It was also recognized that VBA did not possess data sources sufficiently rich in their detail to enable us to do the kind of data comparisons needed to objectively identify potential areas of variance or inconsistency for further investigation.
In response to the GAO findings, we conducted a test in Nashville to learn what factors are critical in the design of consistency measurement tools. That initial test attempted to measure the consistency with which regional offices evaluated three discrete disabilities. The reasons for the selection of these disabilities are as follows. Hearing loss was selected because we anticipated that it would reflect the highest level of consistency across regional offices due to the highly objective nature of the evaluation criteria. Knee disabilities were chosen because they are among our most commonly claimed conditions. While the evaluation criteria for knee disabilities in the VA rating schedule are substantially objective, the application of DeLuca v. Brown, a Veterans Court decision that requires that a rating for limitation of motion take into consideration the degree of additional loss of range of motion due to pain, introduces more subjective judgments into the evaluation. Finally, we chose PTSD because of its high degree of complexity. The Nashville study found that the single most significant factor contributing to rating inconsistency was whether reviewers judged the case “ready to rate.” Where there was agreement that the case was ready to rate, we found a very high degree of consistency between reviewers and original decision makers. Where there was disagreement, consistency was low. Subsequent to these findings, we developed the “ready-to-rate” checklist and mandated its use.
We are also examining data and data sources, including that collected through the RBA 2000 rating-decision-support application, to develop a method for ongoing reviews that would identify possible inconsistencies among regional offices in the award and denial of specific conditions. We have conducted some preliminary data extracts, and we are now working to establish the process, schedule, and support requirements for these reviews.
Over the past year, news articles – particularly those of the Knight-Ridder News Service and the Chicago Sun Times – highlighted the existence of variations in the amount of annual compensation paid to veterans by state. As a result, Secretary Principi requested that the IG conduct a study of possible explanations for the variance. Earlier this year, the IG published its report.
The IG was unable to identify a single causative factor to explain the variance in the amount of compensation paid among differing states. Rather, the IG found a number of factors that directly contribute to variance in compensation payments. These factors include the makeup of the veteran population receiving benefits in each state. The IG reported that veterans with service prior to Vietnam tend to receive less compensation than those from Vietnam and later periods of service. This finding is consistent with anecdotal data from our employees that veterans from earlier periods of service file claims for increased evaluation less frequently than Vietnam Era and later veterans. The IG also found that veterans who are military retirees receive more annual compensation than veterans who did not retire from the military. Similarly, the IG found that veterans who elected to have representation from a national service organization in recent years received over $1,000 more in benefits than unrepresented veterans.
Finally, the IG identified PTSD – its prevalence and its evaluation, including the grant of Individual Unemployability (IU) ratings based on PTSD – as a major contributor to variability in compensation. It is important to understand that in the IG's review of cases, the IG found insufficient evidence to support the grant of service connection for PTSD in some cases; thus, VA’s decision to award benefits was premature. Where VA had not fully developed the cases and verified the “stressors” as required, the IG was not able to validate entitlement to the service-connection award by VA. The IG did not find that these veterans were not eligible, but rather that additional evidentiary development is required to substantiate eligibility.
VBA has conducted its own review of the 2,100 cases reviewed by the IG. Our preliminary findings are that we generally agree with the IG that some of the decisions made were premature. We did, however, find that a large percentage of cases judged to have insufficient development were older cases in which VA statutes prohibit a change in the rating decision. If a condition has been determined to be service-connected for a period of 10 years or more, service connection is protected and may not be severed except for a finding of fraud on the part of the veteran. In other cases, we found that evidence verifying a stressor was of record, but the decision failed to adequately address the evidence. Additionally, in a number of the claims we reviewed, the benefit was granted by the Board of Veterans’ Appeals, and VBA is not authorized to review Board decisions.
VBA has agreed, because of the strong recommendation of the IG, to conduct a review of PTSD claims in which the veteran was awarded a 100 percent disability rating or IU rating in the last five years. In that review, we expect that the majority of the claims will be found sufficient and will not require further development. We also expect to find, based upon our review of the 2,100 IG cases, that further consideration of entitlement in some cases is barred because the benefit granted is now protected by statute.
In some cases, it will be necessary to conduct “stressor verification” development. VA regulations require that, in order to grant service connection for PTSD, there must be "credible supporting evidence that the claimed in-service stressor occurred." Every decision involving the issue of service connection for PTSD claimed to have occurred as a result of combat must include a factual determination as to whether or not the veteran engaged in combat, including the reasons or bases for that finding. Combat status may be determined through the receipt of certain recognized military citations and other supportive evidence. If the evidence establishes that a veteran engaged in combat or was a prisoner of war (POW) and the stressor relates to that experience, the veteran's lay testimony alone may establish an in-service stressor for purposes of service connecting PTSD. In cases in which the stressful event is not linked to combat or POW status, VA will assist the veteran in establishing that the stressful event occurred while the veteran was on active duty and that the veteran was present at the event, including asking the Center for Unit Records Research or the Marine Corps Historical Center to research records that can verify occurrence of the stressor. This is a process that can take up to six months or more.
In addition, before VA can grant service connection for PTSD, VA regulations require a determination as to whether there is medical evidence diagnosing PTSD and linking the veteran's current symptoms to the in-service stressor.
In response to the IG’s recommendation that VA conduct a scientifically-sound study of the major influences on compensation payments, VA’s Office of Policy, Planning, and Preparedness has contracted with the Institute for Defense Analyses to perform a multi-variant analysis of the state-by-state and VA regional office distribution and variation in disability compensation claims, ratings, and monetary benefits to determine if there is a significant correlation to one or more variables. We anticipate receiving an interim briefing in January 2006 and a final report and database by October 2006, and believe that the information obtained will be useful in developing baseline data and metrics for monitoring and managing variances.
Finally, the Office of Policy, Planning, and Preparedness has also undertaken a more detailed analysis to identify differences in claims submission patterns to determine if certain veteran sub-populations, such as World War II veterans or those living in specific locales, have been underserved. We will use the results of this study to perform focused outreach efforts to ensure all veterans have equal access to VA benefits.
VA acknowledges and is concerned that there are variances across the system with respect to average annual benefit payments, and we find it perplexing. We do not, however, agree that “average annual payments” should be the measure by which we judge consistency. Measurement of consistency is complex and cannot be discerned based upon a single measure of state-by-state comparisons of average disability payments.
Annual average payments are not a good way to evaluate consistency in disability compensation awards made in recent years, as these averages include all veterans on VA’s disability compensation rolls, including many whose rating decisions were made in the 1940s and 1950s. Additionally, as the IG found, demographic factors play a role.
We will continue our efforts to better understand this complex and difficult issue, and to identify and reduce inappropriate variability in our decisions. Our objective is to ensure that all regional offices are generating consistently accurate decisions that provide the maximum benefits to which veterans are entitled.
I would now like to turn to the Allen case, as requested by the Subcommittee. In 2001, the United States Court of Appeals for the Federal Circuit interpreted 38 U.S.C. § 1110 as allowing compensation for an alcohol or drug abuse-related disability arising secondarily from a service-connected disability. Allen v. Principi, 237 F.3d 1368, 1370 (Fed. Cir. 2001). VA is concerned that payment of additional compensation based on the abuse of alcohol or drugs is contrary to congressional intent when it mandated in PL 101-508 that benefits not be paid for alcohol or drug related disabilities, and that it is not in veterans’ best interests because it removes an incentive for them to seek treatment for this debilitating compulsion.
The Federal Circuit's interpretation in Allen increases the amount of compensation VA pays for service-connected disabilities. Under the court's interpretation, any veteran with a service-connected disability who abuses alcohol or drugs is potentially eligible for an increased amount of compensation if he or she can offer evidence that the substance abuse is a result of a service-connected disability; that is, that the substance abuse is a way of coping with the pain or loss the disability causes. Under this interpretation, alcohol or drug abuse disabilities that are secondary to either physical or mental disorders are compensable.
VBA’s initial assessment of the ruling in Allen was that the impact would be significant. Therefore, VBA conducted special reviews of a random sample of cases to objectively assess entitlement or potential entitlement to compensation based on the Allen decision.
The first review in June and July 2004 included all rating-related cases otherwise reviewed in our quality assurance review program. The results from this first review indicated that Allen claims received or potential Allen claims identified were associated with mental disorders. As a result, a second review took place from August 16, 2004, to January 14, 2005, limited to mental disorders listed in VA’s rating schedule. A total of 359 cases were reviewed during this period for any potential eligibility under Allen. Thirteen claims for service connection for disability related to substance abuse as secondary to a service connected disability were identified. Potential Allen issues were identified in 29 cases. In 27 of these 29 cases, the condition at issue was alcohol dependence (or abuse) secondary to PTSD or claimed PTSD. Possible reasons for this somewhat limited Allen impact are that abuse symptoms are intertwined with the underlying psychiatric disorder and not easily separated or Allen-type claims are frequently received after veterans are in active treatment. In this regard, in many cases the diagnoses include “history of abuse.” Finally, veterans may not be aware of the potential for compensation under this court ruling.
The Secretary recently transmitted to Congress a draft bill, the “Veterans Programs Improvement Act of 2005.” Section 3 of this draft bill amends sections 1110 and 1131 of title 38, U.S. Code, to clarify that disability compensation benefits may not be paid on account of disease or disability resulting from the abuse of alcohol or drugs, even when the abuse is secondary to a service-connected disability. It also clarifies that an alcohol or drug abuse disability may not be used as evidence of the increased severity of a service-connected disability. Based on our findings from the special reviews we conducted, we revised our cost estimate to indicate a more limited impact.
I want to assure the Subcommittee that we take our responsibility to accurately, fairly, and compassionately decide claims for disability from America’s veterans very seriously. We believe that veterans should get the same result based on the same set of facts regardless of the State in which they reside or the regional official that decides the claim. Rating veterans disability claims is a complex and difficult task, frequently requiring resolution of multiple issues with sometimes conflicting medical evidence. It is a responsibility, however, that we believe can be done competently and compassionately. Our efforts are directed to that end.
Mr. Chairman, this concludes my testimony. I greatly appreciate being here today and look forward to answering your questions.