STATEMENT OF
THE HONORABLE ROBERT H. ROSWELL, M.D.
UNDER SECRETARY FOR HEALTH
DEPARTMENT OF VETERANS AFFAIRS
BEFORE THE
COMMITTEE ON VETERANS AFFAIRS
U.S. HOUSE OF REPRESENTATIVES
ON THE
VETERANS EQUITABLE RESOURCE ALLOCATION (VERA) MODEL

April 30, 2002

Mr. Chairman, it is my pleasure to testify before the Committee on the status of the Veterans Equitable Resource Allocation ( VERA) model.

As you know, VERA was developed at the direction of Congress to replace an outdated historical based allocation system. Over the years, the VERA model has been improved and enhanced to respond in a fair and equitable manner to changes in the practice of medicine and in the delivery of health care services. Proposed changes to the VERA model have been generated from two main sources, internal teams of senior VA health care practitioners, managers, and executives; and external consultants such as the General Accounting Office ( GAO), the RAND Corporation, and PriceWaterhouseCoopers. GAO has been particularly helpful in highlighting areas and challenges that need to be addressed to improve the VERA model. The recommended changes and improvements from outside experts are an excellent endorsement of the effectiveness of the VERA model, because none of them has ever recommended replacing the VERA model. The external experts have all acknowledged that the VERA model is basically meeting its objective of allocating scarce resources in a fair and equitable manner.

This brings me to GAO's most recent report issued in February this year, which is the subject of this hearing. Before I comment on GAO's specific recommendations, I would like to commend GAO for the professionalism and thoughtful analyses that characterize this, their third evaluation of the VERA model. GAO's five recommendations were as follows:

1. better align VERA measures of workload with actual workload served regardless of veteran priority group;
2. incorporate more categories into VERA's case-mix adjustment;
3. update VERA's case-mix weights using the best available data on clinical appropriateness and efficiency;
4. determine in the supplemental funding process the extent to which different factors cause networks to need supplemental resources and take action to address limitations in VERA or other factors that may cause budget shortfalls; and
5. establish a mechanism in the National Reserve Fund to partially offset the cost of networks' complex care patients

VHA is currently evaluating proposed changes to the FY 2003 VERA to be responsive to GAO's recommendations. Final decisions will be made by the Secretary. We hope to have final decisions in time to implement for the FY03 allocation. Some of the issues being addressed are:

• how to address non-service-connected/non-complex care Priority 7 veterans in VERA Basic Vested Care (responds to recommendation 1);
• adjusting the Complex Care and Basic Care price split to reflect actual costs of the two groups (responds to recommendation 3); and
• providing an additional allocation for the very highest cost patients, those whose annual cost exceeds an established threshold (responds to recommendation 5).

I would like to discuss GAO's recommendations.

GAO Recommendation 1 - Better Align VERA Workload Measures

Although inclusion of non-service-connected/non-complex care Priority 7 veterans in the VERA Basic Vested Care category would be a step toward better aligning the VERA allocation model with VA's actual enrollment experience, including these veterans in the VERA model would create financial incentives to seek out more of these veterans instead of veterans with service connected disabilities or those with incomes below the current income threshold or special needs patients (e.g., the homeless), veterans who comprise VA's core health care mission. We experienced uncontrolled growth in the Priority 7 veterans when they were not included in the VERA model, and we do not want to encourage unmanageable growth by including them in the VERA model. Allocation of fixed resources is a zero sum game. Increased resources for Priority 7 veterans would come at the expense of veterans who are service-connected, poor, or who require specialized services. Allocation of resources to areas with a disproportionate percentage of Priority 7 veterans would come at the expense of veterans who live in areas with disproportionately higher numbers of service-connected and lower income veterans. Therefore, we are very carefully weighing how best to address this issue.

GAO Recommendation 3 - Update VERA's Case-mix Weights

GAO has also proposed a change to adjust the price split between Complex Care and Basic Care to reflect the current cost experience between these two groups rather than using a fixed ratio that reflects their FY 1995 relative costs. The Secretary will not approve a change which would create a disincentive for the enrollment and treatment of complex care patients, veterans who need treatment for services such as blind rehabilitation or spinal cord injury.

GAO Recommendation 5 - Establish a Mechanism in the National Reserve Fund

The proposal to provide an additional allocation to networks for the highest cost patients recognizes the impact on those networks with patients whose annual costs exceed and established threshold. These networks would receive an additional allocation equal to the amount that their costs exceeded the threshold. This addresses not only the highest cost Complex Care patients, but also those in the Basic Care group.

GAO Recommendation 2 - More Categories in the VERA Case-mix Adjustment

With regard to recommendation 2, we currently have identified three potential case-mix approaches; however, they affect various networks very differently and we do not yet fully understand these effects. The three potential approaches are:

1. VERA with 44 case-mix categories, as described in the GAO report;
2. VERA with 10 case-mix categories, which is a higher grouping of the 44 case-mix categories; and
3. the Diagnostic Cost Groups (DCGs) with 24 case-mix categories.

Both the first and second approaches contain the foundation building blocks of the current VERA 3 case-mix model. The DCG model is similar to the one used by the Centers for Medicare and Medicaid Services (CMS) for its Medicare + Choice program and is a case-mix model that is based mainly on the diagnosis and demographics of the patient, except in the case of special needs patients, where case-mix is based on utilization factors similar to the VERA model.

While GAO may be correct in recommending more case-mix categories, additional time is needed to evaluate the appropriate method because of the significant differences in allocation results under the three approaches. Therefore, we are considering recommending that the Secretary delay a final decision until FY 2004. Additionally, the RAND Corporation is currently evaluating VERA and will report its conclusions and recommendations this fall. We hope that RAND's analysis will provide information on which a more informed decision can be made on model case-mix adjustment.

The attached table shows the estimated impact on all networks of GAO's recommendations in FY 2002 compared to GAO's report estimates for FY 2001.

GAO Recommendation 4 - Supplemental Funding Process

GAO's fourth recommendation indicates a need to determine why some networks need a VERA adjustment or supplemental allocation, identify factors in the allocation model that create a need for these adjustments, or identify the other factors that may contribute to this situation in some networks. Over the six years that the VERA model has been operational, it has been necessary to make supplemental VERA funding adjustments in four of those years. The supplemental adjustments are intended to assist networks that were unable to operate within their initial VERA workload-based allocations and their locally generated revenues from first- and third-party collections and reimbursements.

Prior to FY 2002, requests for supplemental adjustments would be evaluated in various ways before the Under Secretary for Health made a final decision. The process was not complete until about mid way into the fiscal year. In FY 2002, VHA reengineered the supplemental request process to make the determination part of the initial VERA allocation. This was accomplished by developing updated estimates of each network's projected FY 2002 financial status, to include estimates of all resources that would be available to each network and their estimated expenses for the year. The estimate of available resources included funds carried over from the prior year, estimated collections, estimated reimbursements, and the estimated VERA allocation of the medical care appropriation. The estimated expenses were based on the actual expenses of FY 2001, plus approved budget increases for inflation and pay raises, minus a two-percent efficiency target. Based on this analysis, it was determined that five networks should receive an adjustment to their initial VERA allocation. This adjustment was included as part of the initial VERA allocations on December 7, 2001. The table below provides a summary of VERA adjustments from FY 1999 through FY 2002.

Although we would like to minimize these adjustments by identifying and correcting the causes as GAO recommends, it is also important to evaluate these adjustments in relation to the system-wide impact of the VERA allocation model. The VERA model was used to allocate funds to 22 networks in FY 2002 and required an adjustment of 1.5 percent. It would be unrealistic to expect any model to be 100 percent perfect. However, we need to better understand what is causing certain networks to require adjustments year after year. It is certainly possible that part of the cause may be in the allocation model. However, the difficulty associated with eliminating excess capacity, adjusting the size of the work force, and shifting costly inpatient programs to more efficient health care delivery models in a Federal system may also be contributing factors.

Mr. Chairman, this concludes my statement. I greatly appreciate the opportunity to discuss VHA's progress in improving and refining the VERA methodology. I will be happy to answer any questions the Committee may have.