Provision of VA Long‐Term Services and Supports (LTSS) and Inpatient Care to Veterans Residing in Disadvantaged Areas

Abstract

Area deprivation index (ADI) is an established scale for measuring socioeconomic deprivation at the census tract level. The higher the ADI, the more deprived the area. Previous studies indicate health care use disparities for patients residing in high ADI areas. By using national ADI (range 1‐100), this study examined the association between area deprivation and older Veterans’ utilization of long‐term services and supports (LTSS) provided/ paid by the VA but not Medicare, and hospitalizations provided by both Medicare and VA.Department of Veteran Affairs (VA) and Medicare data for fiscal year 2015 were linked by Veterans’ zip + 4 address in July‐September 2015 (to avoid the snowbird residence effect) with national ADI from the Neighborhood Atlas. Outcomes included receipt of any VA LTSS, nursing home care (NH), personal care services (PCS: Adult Day Health Care, Home Maker/Home Health Aide, Respite), and hospitalizations: provided or purchased by VA, by Medicare and combined. VA medical centers (VAMC) fixed effect models were used to estimate the association of ADI with each outcome, adjusting for Veterans’ predisposing (age, gender, race, urban/ rural residence), enabling (marital status, priority status), and needs (Nosos‐VA cost predictor and JEN Frailty Index) characteristics. Average marginal effects (AME) significant at p < 0.01 are presented for ADI categories 1‐25 (least deprived, reference), 26‐50, 51‐75, and 76‐100 (most deprived).Veterans aged ≥ 66 years using VA services at 141 medical centers merged with ADI data (N = 2.6 million).Groups of ADI 1‐25, 26‐50, 51‐75, and 76‐100 comprised 18.47%, 29.67%, 30.2%, and 21.66% of the population, respectively. Compared to Veterans residing in the least deprived areas, Veterans residing in the most deprived areas and in ADI groups 51‐75 and 26‐50 had 1.01%, 0.56%, and 0.21% higher probability of receiving any VA LTSS, respectively. Veterans residing in areas with ADI 26‐100 had lower probability of receiving any NH care (−0.18%). Veterans residing in areas with ADI 76‐100 and 51‐75 had 0.34% and 0.23% higher probability of receiving any PCS, respectively. Veterans residing in the most deprived areas had 1.16% higher probability of receiving any VA provided/paid hospitalizations, but lower probability of receiving Medicare hospitalizations (−1.21%) and combined VA+Medicare hospitalizations (−0.48%).The association of ADI and hospitalizations in urban areas were similar to the full population with slightly larger magnitude; they were not significant in rural settings except among Veterans with ADI 76‐100 and 51‐75 (−0.77% and −0.53%, respectively). There was no statistically significant difference in the likelihood of VA or Medicare hospitalizations by ADI, but Veterans in the most deprived areas and areas with ADI 26‐50 had 0.3% higher likelihood of VA purchased hospitalizations.The VA appears to target PCS to Veterans residing in highly deprived areas. Hospitalizations, which are Medicare and VA benefits, were less likely to be provided by Medicare in more deprived (primarily urban) areas, but more likely to be provided by VA or purchased by the VA.VA and Medicare should jointly strategize to address the needs of Veterans in more deprived areas.Department of Veterans Affairs.