Declining Access to Alcohol and Drug Abuse Services among Veterans in the General Population

Military Medicine, Mar 2005 by Tessler, Richard, Rosenheck, Robert, Gamache, Gail

Although the Veterans Health Administration has been committed to preserving its capacity to provide specialized substance abuse (SA) services, administrative data from the late 1990s point to reduced access. To explore the possible effects of reduced access to Veterans Affairs services on veterans in the general population, we examined data from the National Household Survey of Drug Abuse from 1994 to 2001. Although the data are not longitudinal at the level of individuals, the annual nature of these surveys makes it possible to chart aggregate changes in clinical needs and service use over time. The results indicate that, whereas veterans' needs for SA services remained high and were consistent across these years, the proportion of veterans in treatment decreased. The proportion of nonveterans receiving SA treatment also decreased but not as steeply as that of veterans. This trend parallels declining delivery of specialized alcohol and drug abuse services within the Veterans Health Administration, although they do not prove a causal relationship.

Introduction

Although federal policy in the middle 1990s sought to preserve the capacity of the Veterans Health Administration (VHA) to provide specialized services for veterans with alcohol and drug addictions, in fact service delivery declined.1-3 This was not an isolated trend. Fiscal crises at the state level also caused cutbacks in the Medicaid program, which led to severe cutbacks in mainstream substance abuse (SA) services. In an effort to maximize the use of available resources, there occurred a shift in emphasis from inpatient care to outpatient care, paralleling larger trends in health care delivery. Despite these efforts, the net result in the VHA system was that the number of veterans who received specialized SA treatment declined and they represented a smaller proportion of all VHA patients.4,5

Surveys and workload data concerning VHA SA programs reveal a substantial reduction in the availability of SA services (especially inpatient beds but also outpatient visits) beginning in the later 1990s.6 Compared with the period 1991-1994, when the number of VHA SA programs increased by almost 20%, the years that followed showed a marked reduction in the part of the budget dedicated to core SA services, with a parallel decline in access to services. The part of the VHA budget dedicated to SA treatment was reduced by 23% between 1997 and 1999, whereas spending for all other services increased by an average of 9%.1 By 1997, the number of inpatient programs had declined by 63%, from 180 programs to 66. Of the inpatient programs that remained, most gave priority to treating SA veterans with comorbid psychiatric disorders.2 In these ways, it appears that budgetary constraints changed the kinds and amounts of SA services the VHA was able to offer.3

This trend is of concern only to the extent that veterans' needs for SA services remained constant during this period. The possibility of declining need is suggested by the aging of the veteran population (which peaked following the Vietnam War) and perhaps also by the cumulative benefits of public health education about SA.

We adopted a general population approach as an alternative to focusing on veterans already in specialized VHA SA treatment programs. This approach requires that SA case identification proceed independently of VHA treatment status. The central questions are whether, and to what extent, there has been a decline in the proportion of veterans with SA disorders in the general population who can receive treatment for those disorders and whether such a decline parallels changes within the VHA.

With this issue in mind, we turned to publicly available data about SA problems and treatment in the general population. The data source of choice, the National Household Survey of Drug Abuse (NHSDA), allowed us to differentiate nationally representative samples of veterans and nonveterans and to examine the interrelationships between need and treatment among them. Although the data are not longitudinal at the level of individuals, the annual nature of these surveys makes it possible to chart changes in clinical needs and service use in the general population of veterans and nonveterans over time.

We formulated the following hypotheses: (1) The prevalence of alcohol and drug abuse among veterans in 1997, 1998, and 2001 would be unchanged relative to the prevalence in 1994, before the reduction in access to VHA SA services. (2) Parallel to the reduction in SA services in the VHA system and elsewhere, there would be a decline in the numbers of both veterans and nonveterans with SA problems who were receiving SA treatment. (3) The rate of decline in treatment received for SA problems would be steeper for veterans than for nonveterans, because of the specific changes in veterans' access to VHA SA services.

Methods

Data Source and Survey Design

The largest and most nationally representative source of information about both the use of illicit drugs and alcohol and the use of treatment services for SA among members of the noninstitutionalized U.S. civilian population is the NHSDA. The NHSDA yields national estimates of the prevalence of SA based on interviews with more than 20,000 persons per survey year. Although the NHSDA is not designed to examine changes in substance use over time for specific individuals, continuity in the methods of sample selection and in the measurement of key variables through study year 1998 provides a reasonable basis for making comparisons across years.