Homeless Veterans Eligible for Medicaid Under the Affordable Care Act

Data were based on the VA Homeless Operations Management and Evaluation System (HOMES). HOMES reflects the primary data collection of all of VA’s specialized homeless programs and offers a source for service providers, policy makers, administrators, and researchers. VA homeless program performance measures indicate that less than 5% of veterans in the HOMES database are ineligible for VA care. This study used client-level data from 114,497 veterans across 142 sites from April 2011 to November 2012.

VA homeless services staff conduct assessment interviews with all veterans who are prospective clients of VA homeless programs to obtain information about sociodemographic characteristics, homelessness and incarceration history, and health status using structured forms. Recent history of homelessness is based on the past month, and history of chronic homelessness is consistent with the federal definition of being continuously homeless for one year or more or having four or more episodes of homelessness in the past three years. Incarceration history is assessed by asking clients how much total time they have spent in jail or prison during their lifetime, which is coded dichotomously as having an incarceration history or not. Medical history is based on self-report of nine medical conditions, and veterans are also asked to rate their general medical health in the past month on a 5-point scale from poor to excellent. Psychiatric history is assessed during the interview and through review of medical records. Veterans are asked whether they have ever been hospitalized for a psychiatric problem.

The total sample was divided into three mutually exclusive groups: those currently covered by Medicaid, those likely to be eligible for the Medicaid expansion coverage, and those not likely to be eligible for the Medicaid expansion coverage. Eligibility for the Medicaid coverage under the expansion was based on age and income requirements. All veterans over the age of 65 were excluded. Variables used to determine eligibility status were reported income, marital status, and number of dependent children. Those who were categorized as likely to become eligible for Medicaid had an income equal to or less than 138% of the federal poverty level in 2012 based on their household size.

Analysis of variance was conducted to compare the three groups on sociodemographic characteristics and health status. With 30 comparisons, a Bonferroni correction was used, and the significance level for all analyses was set at .001 (.05/30). Effect sizes were calculated by using Cohen’s d for continuous variables and differences in percentages for categorical variables, with Cohen’s d values over .5 and differences 5% or larger judged to be notable. Comparisons of groups on general medical and psychiatric conditions were repeated, with control for age, by using logistic regressions, and odds ratios (ORs) were calculated.

Of the 114,497 homeless and at-risk veterans, only 3% reported current Medicaid coverage (Table 1). However, the analysis determined that 78% were likely to become eligible for Medicaid, and 14% were not likely to become eligible. In each of the six states with the largest veteran populations, most were likely to become eligible for Medicaid (71%–87%), and only a small minority was currently covered by Medicaid.

Homeless and at-risk veterans who were likely to become eligible for Medicaid were more likely to be black and chronically homeless than those not likely to become eligible for Medicaid. However, they were less likely to receive VA service-connected disability benefits or to be retired or disabled. [A table in an online data supplement to this report presents results of this analysis.] Compared with veterans who were currently enrolled in Medicaid, those likely to become eligible for Medicaid were more likely to be male and to have a history of incarceration but were less likely to be retired or disabled, had lower total income, and spent more nights in an institution in the previous month.

Overall, homeless and at-risk veterans reported substantial general medical and psychiatric problems, especially substance use disorders (53.7%), chronic pain (36.6%), affective disorders (36.0%), and hepatitis C (13.1%). Those likely to become eligible for Medicaid were less likely than current Medicaid recipients to report having chronic obstructive pulmonary disease (COPD), heart disease, diabetes, chronic pain, affective disorders, and psychotic disorders, but they were more likely to have substance use disorders. [A table in the online supplement presents results of this analysis.] Compared with those likely to become eligible for Medicaid, those not likely to become eligible were more likely to have heart disease, diabetes, posttraumatic stress disorder, and prior psychiatric hospitalizations, but they were less likely to have substance use disorders.

When these comparisons were repeated, with control for age, group differences in substance use disorders were no longer noted, but those likely to become eligible for Medicaid still had better general medical health than current Medicaid recipients. Compared with those likely to become eligible for Medicaid, current Medicaid recipients were more likely to report HIV-AIDS (OR=1.70), COPD (OR=1.51), stroke (OR=1.77), seizures (OR=1.61), and chronic pain (OR=1.54). Compared with those likely to become eligible for Medicaid, those not likely to become eligible were less likely to have hepatitis C (OR=.41) but more likely to have PTSD (OR=2.28).

A small proportion (3%) of homeless and at-risk veterans accessing VA homeless services was currently covered by Medicaid. However, the study found that the large majority (78%) would likely become eligible for Medicaid in 2014 in states that implement the ACA’s Medicaid expansion. This study showed that as a group, homeless and at-risk veterans are a vulnerable population with a variety of general medical and psychiatric problems, especially substance use disorders. Homeless and at-risk veterans likely to become eligible for Medicaid through the expansion were even more likely than other homeless veterans to have substance use disorders. VA provides various forms of substance abuse treatment (outpatient, inpatient, and residential treatment), but Medicaid also covers many of these services in non-VA facilities, suggesting that Medicaid coverage may increase access and choices for the many who can benefit from them.

In states that implement the Medicaid expansion, characteristics of Medicaid enrollees will change as nonelderly, nondisabled adults become eligible for Medicaid. As shown in this study, homeless and at-risk veterans in this newly eligible category may have extensive health needs that either VA or Medicaid-funded providers can address. Having a VA service-connected disability may play a role in whether VA or Medicaid-funded services are used, because those with a service-connected disability may identify with the VA more than other veterans and may be more likely to use VA services (8). As this study found, those likely to be eligible for Medicaid coverage under the expansion were less likely to be receiving VA service-connected disability benefits than those likely to be ineligible for Medicaid coverage.

Thus one potential impact of the Medicaid expansion is increased health care options and better access for many homeless and at-risk veterans (4). However, access to providers who accept Medicaid varies by state and service type, although currently, overall access for Medicaid enrollees is comparable to that of low-income privately insured individuals in the local health care market (9). Education and outreach to homeless veterans may be important because they can be a difficult population to serve and services may not reach those that need them. These issues are important to consider as VA strives to address the needs of homeless veterans (2). VA and the Centers for Medicare and Medicaid Services may benefit from dialogue on how to best provide outreach and coordinated care.

It remains unknown what proportion of homeless and at-risk veterans who will become eligible for Medicaid will actually enroll. The application process may be difficult for homeless individuals who have language and literacy barriers; distrust public bureaucracies; or lack transportation, stable contact information, and documentation (10). Moreover, homeless services, such as provision of housing subsidies, are not covered by Medicaid, and it is currently unclear whether and how other homeless services, such as case management, may be covered (11).

Staff of VA homeless programs can help facilitate Medicaid enrollment among homeless and at-risk veterans and are well positioned to do so. Previous efforts in which VA staff assisted with obtaining non-VA benefits have been successful. For example, a joint benefits outreach project between VA and a colocated Social Security Administration office led to substantial increases in applications for Social Security benefits for homeless veterans, and those who received benefits reported higher quality of life and fewer days homeless (12).

However, there are concerns that enrollment in both VA and Medicaid may lead to fragmented care (3). Lessons learned from veterans dually enrolled in VA and Medicare may illuminate potential advantages and pitfalls of dual enrollment in VA and Medicaid (13,14). It is possible that a small group will be triply enrolled (Medicare, Medicaid, and VA), and this group may need particular attention or service coordination. Thus one recommendation is for VA providers to be more aware of other forms of coverage that their patients may have or are eligible for, such as Medicaid. It may also be important for VA and non-VA providers to communicate about patients and to document this communication in patients’ medical records. At the administrator level, detailed strategic planning between VA and states that decide to implement the ACA’s Medicaid expansion may be needed to advance the nation’s efforts to address the health needs of homeless veterans.

This study had several methodological limitations worth noting. The data were limited to homeless and at-risk veterans who were engaged with VA homeless services and did not include those outside the VA system. Veterans in the sample who were not eligible for VA care could not be identified, although they arguably would be most affected by the Medicaid expansion. Current data on age and income were used to determine who is likely to be Medicaid eligible in the future, and circumstances for some veterans may change. Finally, general medical conditions were based on self-report, which may be unreliable or incomplete. Nonetheless, the results shed light on a population of great public policy concern, which may be influenced by the changing landscape of American health care.

The study estimated that over three-quarters of homeless VA service users would likely become eligible for Medicaid in 2014 in states that implement the Medicaid expansion. Eligible veterans may have various general medical and psychiatric disorders that could be treated by VA or Medicaid-funded services or both. The possible interplay between VA health care and the expansion of Medicaid should be planned for.