VA Health Care - An Analysis Of The Aging Veteran

Transcription

1 AGING VETERANS: WILL THEY OVERWHELM THE VA MEDICAL CARE SYSTEM? by Constance Horgan, Amy Taylor, and Gail Wilensky Prologue: Within the American medical care system, the Veterans Administration (VA) operates a vast subsystem that is charged by law with curing for the health care needs of veterans of the armed services. Outlays-to finance this enterprise of 172 medical centers, forty-nine satellite or independent outpatient clinics, and 195,000 employees consumed some $8 billion in fiscal Even in an era of austerity, the VA has proven resilient politically, handily dispensing with any designs the Reagan administration may have had to check its growth. The future, though, remains less clear. The implications of an aging society have been receiving an increasing amount of attention in the last several years. Although much of the discussion has focused on the viability of Social Security and Medicare s Hospital Insurance Trust Fund, the VA has focused its concern on the aging veteran. Specifically, what impact will the growing population of aging veterans have on the comprehensive service-s offered by the VA medical care system? In its 1978 report to Congress, the Veterans Administration declared: After 1985, the projected veterans demand for VA hospital beds is expected to accelerate rapidly assuming current eligibility and utilization factors. But will the increase be us large us the bulge in the veterans population due to World War II and Korean War might suggest? Three health conomists, Constance Horgan and Amy Taylor of the National Center for Health Services Research and Gail Wilensky, vice-president of Project HOPE s domestic division and director of its Center for Health Information, Research and Analysis, explore this issue. They focus on the extent to which veterans used the VA system in based on data from the National Medical care Expenditure Survey - and are likely to use it in the future.

2 78 HEALTH AFFAIRS Veterans of the U.S. Armed Forces are eligible to use the health services of the Veterans Administration (VA), outlays for which accounted for 3 percent of total U.S. health expenditures in Because eligibility to use the VA system can also extend to veterans otherwise insured by public or private mechanisms, there is reason to ask whether the VA is primarily serving veterans who have other forms of insurance, or primarily those who have no other coverage. This issue of overlapping coverage is particularly important in view of the aging of the veteran population. At age sixty-five, most veterans become eligible for Medicare and receive a higher priority status for use of the VA health care system. Many World War II and Korean War veterans are now approaching or have already reached age sixty-five. Figure 1 shows that while in 1977 approximately 25 percent of adult males over sixty-five years of age were veterans, by 1995 this percentage will have increased to over sixty out of a hundred. Also, as indicated in Figure 2, the number of elderly veterans will more than double between 1980 and 1990, from slightly more than 3 million in 1980 to well over 7.5 million by Although determining the exact number of veterans eligible to use the VA health care system is not straightforward because of the complex criteria used to determine eligibility and priority status for the use of services in the VA system, veterans sixty-five years of age or older can use the VA system for the treatment of other than service-connected conditions without attesting to inability to pay. 2 This change in eligibility exists regardless of Medicare coverage and the, availability of private insurance. It has been argued that the aging of the veteran population will put increasing pressure on the VA health care system. 3 Figure 1 Veterans Over 65 Years As Percent Of All U.S. Males Over 65 Years, Source: A Report on the Aging Veteran, U.S. Senate Committee on Veterans Affairs, p. 23.

3 AGING VETERANS 79 Figure 2 Number Of Veterans Over 65 Years, Source: A Report on the Aging Veteran, U.S. Senate Committee on Veterans Affairs, p. v. The focus of this paper is the extent to which veterans in different age groups and with different types of insurance coverage in fact used the VA system in 1977 and are likely to use it in the future. We base these estimates on the use of three VA services: short-term hospital inpatient care, ambulatory physician services, and prescribed medicine use. Also, we have limited this examination of health care utilization patterns of veterans to the probability that a veteran will have used the VA health care system in the course of a year for the three specified services; but we do not discuss the amount of use of these services. The emphasis of this analysis, therefore, is not on producing national estimates of use of the VA health care system but rather on issues related to which veterans use the VA and why, and what impact this is likely to have on the VA health care system in the future. The analysis is based on data from the National Medical Care Expenditure Survey (NMCES), household survey of the civilian noninstitutionalized population in NMCES was funded by the National Center for Health Services Research and cosponsored by the National Center for Health Statistics. About 14,000 randomly selected households were interviewed six times over an eighteen-month period about use, expenditures, and insurance coverage for health care in calendar year Standard demographic and socioeconomic data were collected as well as information on insurance coverage and whether any household members had ever served in the Armed Forces or had a service-connected disability, reflecting the veteran status of all individuals in the survey. Information was also collected on medical events occurring in Events for which the VA was reported as the source of payment were assumed to have occurred in the VA health care system. This definition therefore includes both care which took place in a VA facility as well as care for which the VA may have paid for services rendered at a non-va facility for eligible

4 80 HEALTH AFFAIRS veterans. Our analysis is limited to male veterans twenty years of age or older, as the number of female veterans was too small for separate analysis. Finally, persons residing in nursing homes or other institutionalized settings are not included. The institutionalized population represents a relatively small group of VA users, although their importance to the VA with respect to level of use is likely to be disproportionate to their numbers. In the first section of the paper we will look at survey findings concerning the public and private insurance coverage of the veteran population and then at the effects of age and insurance on use of the VA health care system. Finally, we discuss the implications of the aging of the veteran population for the likelihood of use of the VA system in the future. Insurance Coverage And Use Of Selected VA Health Care Services In 1977, about 4 percent of the approximately 29 million male veterans in the United States used some VA health services (ambulatory physician visits, hospital stays, or prescribed medicines; Table 1). 5 The characteristics of these veterans have been described on the basis of NMCES data. 6 Veterans using the VA in 1977 appeared to be different from veterans who received their health care elsewhere. They were more likely to be poor, elderly, and have had less education. They also appeared to be in poorer health, and were much more likely to have a service-connected disability and more likely to report a mental condition. These differences are not unexpected, given VA eligibility criteria with respect to serviceconnected disabilities, age, and ability to pay. These observations are borne out by the relationship between insurance coverage and use of the VA system shown in Table 1. 7 In fact, most veterans were covered by private insurance in 1977 and these were least Table 1 Likelihood of Use of Selected Health Service by Veterans According to Insurance Coverage and Source of Care (row percents) a Use of services Veterans without Total Any V.A. No V.A. use veterans Percent distribution (1000s) All veterans ,859 Private insurance ,686 Medicare Medicaid 7.3* Uninsured ,514 a Selected services include ambulatory physician services, hospital stays, and prescribed medicines. *The relative standard error of this estimate exceeds 30 percent and is thus too high to permit confidence. Source: National Medical Expenditures Survey. National Center for Health Services Research.

5 AGING VETERANS 81 likely to have used VA services. Veterans with Medicare coverage were four times as likely to have used the VA as those with private insurance, and a comparable but threefold difference is found for veterans without any insurance, although the latter group were much less likely to have used any health services at all. 8 However, when they used services, they were more likely to have used those provided by the VA than the privately insured. A similar picture emerges if we consider only those veterans who used the VA system. Among all users of VA services, almost half were uninsured or on Medicare (see Table 2), while the corresponding percentage for those using services outside the VA system only was less than 20 percent. Even more important in view of the aging of the veteran population is the strong relationship between age and use of the VA (see Table 3). Veterans below the age of forty-nine were half as likely to use VA services as those above age fifty. In order to test whether the overall relationships between age, insurance status, and use of the VA are independent of other factors which might influence use of the VA, we examined a regression equation ex- Table 2 Likelihood of Use of Selected Health Service by Veterans According to Insurance Coverage and Source of Care (column percents) a Total veterans (in thousands) Percent with private insurance Percent with Medicare Percent with Medicaid Percent Uninsured Total Use of services Veterans without Any V.A. No V. A. use 1,103 19,528 8, % 80.7% 77.1% * a Selected services include ambulatory physician services. hospital stays, and prescribed medicincs. *The relative standard error of this estimate exceeds 30 percent and is thus too high to permit confidence. Source: National Medical Expenditures Survey, National Center for Health Services Research. Table 3 Percent of Veterans a Using the VA for Selected Health Services b by Age, United States, 1977 Age in years Percent of veterans using the VA a The table excludes veterans who used no medical services in b Selectcd services include ambulatory physician services, hospital stays, and prescribed medicines. Source: National Medical Expenditures Survey. National Center for Health Serviccs Research

6 82 HEALTH AFFAIRS plaining the use of VA services by controlling for the characteristics of the individual veteran who used any medical services in This equation expresses the probability of using the VA as a function of insurance coverage, presence of a service-connected disability, access to a VA facility, age, income, health status, and other sociodemographic variables. 9 As expected in view of the eligibility criteria mentioned above, the most important characteristic explaining use of the VA system was the presence of a service-connected disability. Having private insurance lowered the probability of use, as observed in Table 2. Nonwhites were more likely to use the VA than whites, as were persons living in the Northeast compared to other regions of the U.S. Health status clearly played a role, in that veterans with a higher number of disability days in 1977 and veterans in only fair or poor health were more likely to use the VA than others. Persons employed throughout 1977, and those who never worked in 1977 (most of whom are elderly), were less likely to use the VA than the unemployed. Finally, in comparison with other factors such as lack of private insurance or poor health status, being over sixty-five did not itself increase the likelihood of use of the VA system. This is not to say that age is an unimportant factor in predicting the impact of the aging of the veteran population on the VA health care system. As can be seen in Table 4, in 1977 the size of the near-elderly veteran population, years, was close to four times that of veterans over sixty-five years of age. More importantly, only 27 percent of elderly males were veterans in 1977, while 45 percent of the cohort in 1977 were veterans, and as much as 75 percent in the year cohort. The bulk of the elderly male population for this decade and into the 1990s are the veterans of World War II and the Korean conflict. Even taking into account normal attrition due to death, there clearly is a large increase in the number of veterans who will turn sixty-five and become eligible to use the VA system in this decade. Table 4 Adult Male Population According to Veteran Status, United States, Total Percent Age in years (In 1000s) Veteran Other , I 1, , , , , , Source: National Medical Expenditures Survey, National Center for Health Services Research.

7 The Implications Of An Aging Veteran Population AGING VETERANS 83 The important question is how many of these elderly veterans will survive and choose to use VA health services. As pointed out before, most elderly veterans are covered by Medicare. In addition, about twothirds are also covered by private insurance supplementary to Medicare 10. First, in order to ascertain whether the influence of private insurance coverage on VA use observed in Tables 2 and 3 also holds for the elderly population, thus reducing the additional burden that aging veteran population will place on the VA health care system, we used regression analysis to predict use of the VA for the three specified services by the elderly veteran. We found that having private insurance in addition to Medicare in fact lowers the likelihood of VA use, while service-connected disabilities and poor health status increase this likelihood, as was found for the veteran population as a whole; Second, we aged our base population (that is, the 29 million veterans in our survey) from 1977 to 1982 and Data on age, service connected disability, health status, education, race, region of residence, and presence of a VA facility came from these aged files. Data on percent married and SMSA residence were calculated from census data, and future physician/ population ratios were obtained from the Bureau of Health Professions. It was assumed that the percent of elderly veterans in poverty and the percent employed, as well as disability days and region of residence remain the same. The likelihood of having private insurance and using the VA was predicted using two regression equations based on NMCES data. These indicated that the probability of having only Medicare coverage and of using the VA in the future remains about-the same for the elderly veteran as in Combining these probabilities with the number of elderly veterans predicted for 1987 indicated that although the likelihood of use of the VA system for ambulatory physician visits, short-term hospital stays, and prescribed medicines remains the same, the increase in the absolute number of elderly veterans means that more than twice as many elderly will use the VA system for the services specified between 1977 and However, the eventual impact on the VA system as a whole is probably smaller than suggested by the increase in the elderly veteran user population projected for the decade of the eighties. Although the number of new elderly users of VA services will more than double by 1987 due to the increase in the aged veteran population, overall use of the VA system will not increase by nearly as much because elderly users will remain the minority of total users. As shown in Table 5, based on projections by the VA up to 1995, the percentage of veterans over age sixtyfive will increase to 28.6 percent, while the total number of veterans will not increase. Our calculations indicate that total users of the VA system

8 84 HEALTH AFFAIRS Table 5 Projected Increases in the Population of Elderly Veterans, Total veterans (in 1,000s) 29,459 29,827 29,699 29,090 28,132 Percent 65 and over Percent less than Source: Adapted from A Report on the Aging Veteran U.S. Senate Committee on Veterans Affairs, p. 19. for the services examined would increase by about 16 percent if the number of elderly users doubled. 13 In addition, projections of the change in the nonaged veteran population suggest that the number of users of the VA system under the age of sixty-five will decline by 15 percent over this time period. 14 This decline can be expected to offset at least part of the increase in use by the elderly veteran population, although this assumption does not take into account the distinction between the number of users and the potential extent of use by the elderly, who are likely to be heavy users of services because of the decline in health status with age. It should also be recalled that our data excluded use of long-term care. 15 The potential increase in need for long-term care services which accompanies an aging population is important to the VA. Private insurance and Medicare coverage for nursing home care is minimal, and our data show that insurance coverage is an important determinant of VA use for the three services examined. Given the lack of coverage for nursing home care, the VA may well experience a larger increase in demand for this type of service. In summary, this analysis suggests that the increase in use of the VA health care system for the three services examined is not likely to be as large as one might expect given the magnitude of the projected increase in the elderly veteran population, despite their change in eligibility status at age sixty-five. We have shown that the holding of private insurance coverage lowers the use of the VA system. Thus, increased pressure on the VA system may depend more on whether these veterans have private insurance in addition to Medicare than on the mere increase in the size of this population.

9 AGING VETERANS 85 NOTES 1. H. Sapolsky and S. Wallach, The Veterans Health Care System in Federal Health Programs, ed. S. Altman and H. Sapolsky (Lexington Mass.: Lexington Books, 1981). 2. Veterans who have a service-connected disability have the highest priority for receiving care through the VA; veterans who state under oath that they are unable to pay for necessary hospital care are also eligible for treatment. 3. U.S. Senate Committee on veterans Affairs, A Report on the Aging Veteran, (Washington, D.C.: Government Printing Office, 1978). 4. For the interview instruments, see G.S. Bonham and L.S. Corder, National Health Care Expenditures Study, Instruments and Procedures 1: NMCES Household Interview Instruments, DHHS Publication No. (PHS) (Washington, D.C.: U.S. Department of Health and Human Services Research, 1981); and for a detailed description of the sample design, see S.B. Cohen and W.D. Kalsbeek, National Health care Expenditures Study, Instruments and Procedures 2: NMCES Estimation and Sampling Variances in the Household Survey, DHHS Publication No. (PHS) (Washington, D.C.: U.S. Department of Health and Human Services, National Center for Health Services Research, 1981). 5. Unpublished data from the VA indicate that approximately 10 percent of veterans used the VA health care system in This VA estimate is based on patient encounter records which the VA maintains and reflects total use of the VA health care system. The NMCES estimates of 4 percent reflects a partial picture of VA use, as noted previously. Therefore, a lower NMCES estimate was expected. The exclusions from the NMCES estimate fall into two categories. First, a certain amount of care provided in the VA is for veterans who for the purpose of NMCES would be classified as institutionalized. Included in the VA estimate is the use of long-term care facilities, domiciliaries, and psychiatric hospitals. Additionally, many beds in short-term VA hospitals are actually used for extended care. Second, a large amount of outpatient care in the VA is provided by nonphysicians. Such providers include physical therapists. social workers, pharmacists, mental health counselors, dentists, optometrists, and others. The NMCES estimate includes only visits to physicians. 6. Constance Horgan, Amy Bernstein, and Marc Berk, Use of Health Care Services by Veterans, Paper presented at annual meeting of the American Public Health Association, The insurance categories have been defined to be mutually exclusive. Medicare includes Medicare coverage only or Medicare combined with private practice or other insurance. Medicaid includes veterans who had any Medicaid coverage, but no Medicare. Uninsured refers to those veterans who did not have insurance coverage for all of If a veteran was uninsured for only part of 1977, he would have been classified according to the type of coverage he had when insured. Private refers to all other coverage, mainly private insurance, but also including CHAMPUS and CHAMPVA. 8. Of users of the VA health care system, 28 percent were covered by Medicare but only 16 percent were sixty-five years or older, suggesting that some veterans who use the VA are covered by Medicare because of disabilities rather than age. 9. The actual estimated equations are not presented in this paper. Results are available from the authors on request. The findings are discussed in terms of their statistical significance or lack of it. 10. G.C. Cafferata. Private Health Insurance Coverage, NMCES Report Series, National Center for Health Services Research, Department of Health and Human Services, forthcoming. 11. The data were aged by adding five and ten years, respectively, to a veteran s age in Life tables were used to estimate the surviving population. The means of the relevant variables were then calculated from these aged data files.

10 86 HEALTH AFFAIRS 12. Our estimates indicate that the number of elderly veterans will increase by almost three million from 1977 to The estimated equation predicts that 7 percent of these veterans will use the VA for the selected health care services. This means that the number of elderly users will increase by about 200, This estimate is based on data presented in Horgan, Bernstein, and Berk, See note The projected number of veterans age for 1982 and 1987 was obtained from unpublished VA data. A weighted average of the probability of using the VA for the three specified services was calculated using five-year age groupings based on NMCES data. 15. In 1982, 58 percent of beds in VA Medical Centers were devoted to medical/surgical beds. The balance of beds were psychiatric, domiciliary, or nursing home (VA Annual Report, 1982). An earlier version of this paper was prepared for the 95th annual meeting of the American Economic Association, New York, New York, December, We are grateful to Sandy Smoot and Karen Pinkston of Social & Scientific Systems, Inc., for their programming assistance and Renate Wilson for her helpful comments. The views expressed in this paper are those of the authors and no official endorsement by the National Center for Health Services Research, U.S. Department of Health and Human Services or Project HOPE is intended or should be inferred.