For any disability compensation program, three important measures of performance in claims processing are timeliness, accuracy, and consistency. They are also a focus of criticism for VA disability compensation, which we discuss in this section.
6.3.1.1 Timeliness
Compared to the other disability programs, VA performance in terms of timeliness is poor. The average time for VA to complete a claim (without appeals) in FY2006 was 177 days [36].55 In comparison, the average for SSDI was 88 days in FY2006 [37], and OPM staff report that the FERS/CSRS average is currently 38 days.
Although we were not able to obtain an average for the FECA program, we do have some related measures of timeliness. For FECA in 2002, 96 percent of claims for trau- matic injury were adjudicated within 45 days, 91 percent of claims for basic occupa- tional disease were adjudicated within 90 days, and 76 percent of claims for extensive occupational disease were adjudicated within 180 days [38].
Averages were also not available for the DES program, but we were able to obtain these results for FY2005 from [33]:
54. A recent report [6] has also recommended that raters in the VA program have better ac- cess to medical expertise, based on the fact that their decisions require understanding medical evidence and sometimes “weighing conflicting medical opinions.” That report also makes the point that the other disability programs make more use of medical exper- tise in reaching disability decisions than does VA. Because that report describes in detail the role of health care professionals in the various disability compensation programs, that information is not included here.
55. This refers to the time between VA’s receipt of the claim and the veteran’s notification of VA’s decision. The figure 177 days represents the average for “rating-related actions” for compensation and pension combined.
• Army (both MEB and PEB): 13 percent of claims were completed within 30 days, 32 percent in 31-60 days, 35 percent in 61-120 days, and 21 percent in 120 days or more.
• Navy (PEB only): 58 percent of claims were completed within 30 days, 26 percent in 31-60 days, 9 percent in 61-120 days, and 6 percent in 120 days or more.
• Air Force (PEB only): 53 percent of claims were completed within 30 days, 22 percent in 31-60 days, 15 percent in 61-120 days, and 9 percent in 120 days or more.
Because of the differences across programs in the work required to process a claim, it is difficult to say whether VA’s timeliness problems are due to the complex nature of its disability decisions, staffing shortages, low productivity, or some other factors. To know how best to address its problems with timeliness, it would be useful for VA first to dis- aggregate that 177-day average so that it understands what stages of the claims process are contributing most to the total processing time.
With respect to specific strategies to improve timeliness, VA already does make use of “Tiger Teams” to deal with cases that are designated as high priority at any given time, such as very long-standing cases or cases where the veteran is very old or terminally ill. Because the success of those teams comes from the fact that they are made up of the most experienced staff, unfortunately the Tiger Team approach is not something that VA can replicate on a larger scale (i.e., there are not enough experienced employees to staff a large number of Tiger Teams). VA might also learn from SSA’s new Quick Disability Determination (QDD) process, which involves using a predictive model to identify cases with a relatively high probability of being granted benefits and then try- ing to act on those cases within 20 days [39].
6.3.1.2 Accuracy
Accuracy is another major dimension of the quality of claims processing, and VA’s ac- curacy rate in 2006 was 88 percent [36].56 VBA measures accuracy using its Systematic Technical Accuracy Review (STAR) program, which determines accuracy at the na- tional and RO levels based on the review of at least 120 claims per RO annually. VA’s accuracy rate seems low, especially in comparison with the overall accuracy rate for
56. This rate is based on whether all issues in the claim were addressed, whether the claim was developed in compliance with the Veterans Claims Assistance Act, and whether the rating decision, effective date, and payment date were correct.
SSDI, which is 96 percent [40]. (We were not able to obtain overall accuracy rates for the other programs.) However, when comparing VA’s accuracy to SSDI’s, it is impor- tant to recall that there are differences between the programs in their requirements for processing a claim. In particular, the fact that VA has to rate the severity of a disability creates more potential for error than the yes-or-no disability decision that is required for SSDI.
Regarding specific quality review practices in other programs, SSDI has two types of quality reviews. In its “pre-effectuation review,” SSA uses a profiling system to identify the most error-prone types of claims, and then 50 percent of those claims are selected for review [41]. There is also a separate, smaller quality review effort that reviews a random sample of cases from each DDS.
In the program for FERS and CSRS, OPM staff report that their official quality review process includes drawing a random sample of all claims (both disability and non- disability) and distinguishing between substantive errors (affecting monetary pay- ments) and procedural errors. However, that process analyzes accuracy only for all claims and does not have results available just for disability cases.
For FECA claims, OWCP does not measure system-wide accuracy [38]. Instead, OWCP staff say that some individual DOs use accuracy as an explicit performance measure for individual employees.
In the DES, each service has its own approach to quality review. The Army takes a monthly 30-percent sample of cases for review, and there are also mandatory reviews of certain pre-specified types of cases. The Navy’s approach consists of sending the case to the senior medical officer on the Council of Review Boards. The Air Force has no for- mal accuracy measure or accuracy review process.
In comparing other programs’ practices with VA’s, the only practice that is substan- tively different from VA’s is SSA’s focus on the most error-prone cases. Incorporating this element would require expanding VA’s STAR program, but it could be worthwhile because it would result in a disproportionately large gain in accuracy for any given in- crease in the number of reviews.
6.3.1.3 Consistency
Measuring consistency in disability programs is difficult, and none of the programs currently has the components that GAO recommends for that task: “(1) the use of multivariate regression analysis examining disability decisions along with controlling factors to determine whether the decisions are consistent and (2) an in depth inde-
pendent review of a statistically valid group of case files to determine what factors may contribute to inconsistencies” [33].
VA currently has no measure of consistency [42], although a study that includes rec- ommendations for ways to improve consistency has recently been completed [43]. One problem is that the STAR data do not contain a sufficient number of cases to deter- mine consistency by type of disability across ROs. One of VA’s strategies is to “promote consistency through training and communication” [44], although it appears that that approach is not sufficient. Two studies have documented variation across states in av- erage disability compensation payments, and the state-level differences cannot all be attributed to state-level differences in the characteristics of the population of disabled veterans [43, 44]. In addition, a survey by an NVSO of its National Service Officers in November 2003 indicated that respondents thought there were consistency problems across ROs [46]. Recommendations for improving consistency include standardizing raters’ training, increasing the standardization of medical examinations, increasing the number of claims reviewed, consolidating the rating process into fewer locations, and developing metrics for monitoring consistency [43].
There have also been criticisms of SSA regarding consistency. One source of inconsis- tency across SSA claims examiners seems to be simply the nature of the task of deter- mining medical eligibility. In an evaluation of SSA’s quality assurance processes, [41] found that “many people ... believe that expert adjudicators could reasonably disagree on medical eligibility for an unknown number of cases because of subjective factors. Some say the share of such ‘close call’ cases is almost zero, but others say it could be as high as 20 percent; most suggested 5 to 10 percent.” Consistency across locations (spe- cifically DDSs) is also a concern [47], although an element of SSA’s new DSI process should help address it. As described previously, the DSI changes include moving the first appeal from the state level to the national level. The centralization of decision- making at the national level earlier in the appeals process is expected to help ensure consistency.
Among the other programs, FECA and DES do not measure or analyze consistency sys- tem-wide, although Navy staff indicated that having only one location for processing of Navy claims does help ensure consistency. The FERS/CSRS program does not need to be concerned with inconsistency across locations because it has only one location. Clearly, physical consolidation is viewed as a way to reduce inconsistency in disability programs, and it is an approach that is already being considered by VA, as discussed in the next section.