Fewer Potentially Avoidable Health Care Events in Rural Veterans with Self-Directed Care versus Other Personal Care Services

Abstract

Background:

Rural residents face more barriers to healthcare access, including challenges in receiving home- and community-based long-term services, compared to urban residents. Self-directed services provide flexibility and choice in care options and may be particularly well suited to help older adults with multiple chronic conditions and functional limitations who reside in rural areas remain independent and live in the community.

Methods:

We conducted a retrospective observational study to understand whether differences in health outcomes between Veteran-Directed Care (VDC), a self-directed Veterans Health Administration (VHA)-paid care program, and other VHA-paid home- and community-based personal care services vary in rural/urban location. The sample included 37,395 Veterans receiving VHA-paid home- and community-based long-term care services in FY17. Our primary outcomes were changes in monthly incidence of VHA or VHA-paid community acute care admissions, nursing home admissions, and emergency department (ED) visits. We used fixed effects logistic regression models on unmatched and coarsened exact matched cohorts, stratified by rural/urban location.

Results:

Both urban and rural VDC recipients were significantly less likely to be admitted to VHA-paid nursing homes, compared to those receiving other VHA-paid personal care services (rural: incremental effect = −0.22, [ −0.30, −0.14]; urban: incremental effect = −0.14, [−0.20, −0.07]). Rural, but not urban, VDC enrollees had significantly fewer VHA-paid acute care admissions and ED visits, relative to recipients of other VHA-paid personal care services (acute care, rural: incremental effect = −0.07, 95% CI = [−0.14, −0.01], urban: incremental effect = −0.01, [−0.06, 0.03]; ED, rural: incremental effect = −0.08, [−0.14, −0.02], urban: incremental effect = 0.01, [−0.03, 0.05]).

Conclusions:

VDC recipients had fewer incidents of potentially avoidable VHA-paid health care use, compared to similar Veterans receiving other VHA-paid personal care services. These differences were more pronounced among rural VDC recipients than urban VDC recipients.

Introduction

Residents living in rural areas face more challenges in accessing care compared to those living in urban areas.^1–3 The challenges are related to multiple supply- and demand-side factors. For example, rural residents are more likely than urban residents to experience lack of high-quality care, and have access to few healthcare facilities.^2,3 Compared to urban residents, rural residents have reduced access to providers, including paid home caregivers, and to medical transportation services.^1–3 Rural residents also experience greater communication barriers with providers due to less advanced information technology and have longer driving distances to care providers than urban residents.^1–3 In addition, rural residents have relatively lower incomes and face greater financial constraints compared to urban residents.² Older adults with functional and cognitive limitations living in rural areas experience more barriers to home- and community-based long-term services, are less likely to be discharged from nursing homes, and have higher mortality rates, compared to similar urban residents.^4–6

Unmet needs for long-term care services and supports are of special concern because they can lead to adverse outcomes, such as falls and injuries, increased limitations in activities of daily living (ADLs), and increased frequency of emergency department (ED) visits.⁷ Understanding the best methods for providing long-term care in rural areas may help address unmet needs for care and reduce the risk of adverse outcomes for older adults living in those areas.^7–8

Self-directed services may be particularly well suited to older adults living in rural areas because they provide flexibility and choice in care options. Self-directed service programs provide enrollees with options to determine for themselves the services and supports they need to remain independent and live in the community. Previous research established that self-directed services are associated with fewer unmet needs for long-term care, improved patient satisfaction, and lower risk of adverse outcomes, including injuries, hospital readmissions, and ED visits, compared to other home- and community- based long-term services.^10–14 Participants’ ability to hire family members and neighbors as paid caregivers through self-directed services may be a particularly effective way to surmount access challenges in rural areas.⁹ However, there is limited evidence about the relative effectiveness of self-directed service programs among rural and urban older adults needing long-term services and supports.

Understanding the relative benefits of self-directed services vs. other paid home- and community-based personal care services in rural areas is of particular interest in the Veterans Health Administration (VHA). Twenty percent of older adults in the US are Veterans; about 3 million Veterans are aged 65 years and older and living in rural areas.¹⁵ With increasing needs for long-term care services among older Veterans, the VHA has adopted self-directed models in the design and implementation of home- and community-based long-term services and supports.¹⁵ Self-directed service programs, such as VHA’s Office of Geriatrics & Extended Care (GEC) Veteran-Directed Care (VDC) program, facilitate coverage for Veterans living in areas with limited supply of agency care and personal care aides and help direct resources and spending to services that are of highest perceived value to Veterans.^15–19

The goal of the VDC program is to reduce the need for nursing home placement of Veterans with multiple chronic conditions and functional or cognitive limitations.¹⁸ Our prior work found that participants in the VDC program and other VHA-paid personal care services programs (i.e., homemaker/home health aide [H/HHA], home respite, and contract adult day health care [CADHC]) have similar hospitalization and cost trajectories from before to after enrollment.²⁰ However, the degree to which VDC participants have different trajectories of VHA nursing home and ED use compared to recipients of other VHA-paid personal care services programs is unknown. The extent to which potential effects of the VDC program differ by rural and urban location is also unclear. Moreover, the VDC program is not operational at all VAMCs, and there are substantial variations in the proportions of Veterans enrolled in VDC across VAMCs.^{17, 18} Consequently, further analyses can inform decision-making regarding targeting resources and efforts to expand self-directed service programs both within VHA and in more general settings.

The objective of this study is to understand whether rural and urban Veterans differentially had less VHA or VHA-paid community nursing home, acute care or ED use following enrollment in VDC compared to recipients of other VHA-paid personal care services, namely H/HHA, home respite, and CADHC. In analyses stratified by rural or urban Veteran location, we compare hospital, nursing home, and ED utilization in the 12 months before and after service initiation for Veterans receiving VDC and other VHA-paid personal care services.

Methods

Sample and Study Setting

Our sample included Veterans who were enrolled in VDC, and the comparison groups included Veterans who were enrolled in other VHA-paid home- and community-based personal care services: H/HHA, home respite, and CADHC. All of these VHA-paid personal care services programs seek to reduce unnecessary nursing home placement of Veterans with multiple chronic conditions and functional limitations. In contrast to the other VHA-paid personal care services programs, VDC provides monthly budgets and long-term services and supports options counseling for enrollees to manage their purchases of equipment and services that help them continue living in the community.⁹ Eligible enrollees in the VDC program include Veterans who need over 20 hours of home care per week, are eligible for nursing home admissions, or have three or more ADL dependencies or significant cognitive impairment.²⁰ The VAMCs have partnered with Aging and Disability Network Agencies (ADNAs) to offer the VDC program, and the referring criteria are flexible across VAMCs.^18–20

We compared Veterans who enrolled in the VDC program (with or without enrollment in other VHA-paid personal care services) to Veterans who enrolled in at least one other VHA-paid personal care services program in FY17. For Veterans receiving VDC, the first VDC service date was defined as the initiation of service. For Veterans in the comparison groups, the first VHA-paid personal care service date in FY17 was identified as the service initiation date. For every enrollee, we examined a baseline period of 12 months prior to service initiation and an outcome period of 12 months after service initiation; each month was defined as a 30-day time period. Veterans who had at least one VHA inpatient or outpatient visit in the baseline period were eligible to be included in our sample. Veterans without a residential address in the records were excluded from the sample due to missing rural/urban location information (n = 26).

As in our previous examination of hospital use among VDC participants²⁰, we included 38 VAMCs in the study. Out of the 38 VAMCs, 21 sites had active VDC programs in FY17 and 17 sites had an interest in implementing the program but did not have active VDC programs in FY17.²⁰ The definition of an active VDC program was enrolling at least 5 Veterans per month for the majority of the 12 months in FY17. Veterans in the comparison group included other VHA-paid personal care service enrollees from the sites with active VDC programs to control for facility-level variations in patterns of care and from the sites without active VDC programs to deal with individual-level selection bias, such as unobserved medical conditions that relate to both the receipt of VDC and health care use.²⁰

Data Sources and Study Variables

We used inpatient, outpatient, VHA-paid care and vital statistics data from the VHA Corporate Data Warehouse. The outcomes were VHA or VHA-paid community hospital use (incidence of any all-cause VHA and VHA-paid community hospital admissions per month), nursing home use (incidence of any VHA and VHA-paid community nursing home admissions per month), and Emergency Department (ED) use (incidence of any VHA and VHA-paid community ED visits per month). The nursing home use data included the use of VHA community living centers, VHA-paid contract nursing homes, and State Veteran Homes. We stratified our analyses by whether the Veteran lived in a rural or urban area at the time of service initiation. We used the rural/urban location information derived from the Rural-Urban Commuting Area (RUCA) codes based on Veterans’ residential addresses. Urban location included RUCA codes of 1.01 or 1.1; rural or highly rural locations with RUCA codes other than 1.01 or 1.1 were coded as “rural”.¹⁷ Finer gradations of rurality were precluded by the size of our VDC sample.

Other explanatory variables included enrollment in the VDC program, demographics such as age, gender and race, number of Elixhauser comorbidities²¹, spinal cord injury (SCI), VHA enrollment priority group 1A (indicating service-connected disability), Medicaid eligibility, receiving Aid and Attendance benefits designed to help eligible Veterans pay for assistance with daily needs, traumatic brain injury (TBI), dementia, NOSOS risk score (risk adjustment for VHA patients)²² and Care Assessment Needs (CAN) score²³ indicating the risk of one-year mortality.

Statistical Analysis

Using VHA administrative data on health status and healthcare use, we compared the difference in outcomes from pre- to post- initial enrollment in the VDC program with the difference in outcomes from pre- to post- initial enrollment in other VHA-paid personal care services programs in FY17 for Veterans living in rural or urban areas. We used logistic regression models stratified by rural/urban location to estimate the relationships between VDC receipt and utilization of hospital, nursing home, and ED services. Our primary analyses included an indicator for enrollment in the VDC program after the index date, an indicator for enrollment in other VHA-paid personal care services at inactive sites after the index date, facility fixed effects and time (in months) fixed effects. We calculated the average incremental effects of the treatment (enrollment in VDC after the index date) on the treated. STATA (Version 14.1) was used to conduct the analyses.

In sensitivity analyses, Veterans were matched on time-invariant covariates (Elixhauser comorbidities, age, dementia, SCI and TBI) using the coarsened exact match method to create groups of Veterans that had similar baseline characteristics and health status but were different in the VHA-paid care services they received. Additional sensitivity analyses included using a different definition of the post-enrollment period (setting 2 months after the initial receipt of VHA-paid care programs as the service initiation date); excluding VDC enrollees who did not receive any other VHA-paid care services prior to the initiation of VDC service; excluding decedents with death dates in the outcome period; repeating the analyses on subsets of Veterans who received at least one service per month for at least 3 or 6 months in the outcome period; and excluding Veterans with SCI.¹⁷ As a robustness check, we added a series of models that included hospice admission incidence as the outcome — we would not expect hospice admissions to decrease as a result of the receipt of VDC because the use of hospice services when needed is considered an indicator of improved access.²⁴

In another sensitivity analysis, we calculated contrast between two marginal effects. To do this, we included an interaction term between an indicator for group membership (VDC or comparison) and an indicator for the post-enrollment period in our models. We calculated the contrast in predicted differences in outcomes from pre- to post-enrollment in VHA-paid care between the VDC group and comparison groups.

The study was approved by VA Boston IRB (protocol #3169).

Results

The VDC group included 962 Veterans (387 in rural areas and 575 in urban areas), the comparison group at sites with VDC included 21,100 Veterans (7,245 in rural areas and 13,855 in urban areas) and the comparison group at sites without VDC included 15,319 Veterans (6,099 in rural areas and 9,220 in urban areas).

Among Veterans in both rural and urban groups, based on standardized mean differences of 10.0 or higher, VDC recipients had higher enrollment priority status, were more likely to receive aid and attendance or have sustained a spinal cord injury, compared to recipients of other VHA-paid personal care services programs. ²⁵ Veterans in the VDC group were younger but had higher NOSOS risk scores compared to those in the comparison groups (Table 1). For rural VDC enrollees, the incidence of VHA or VHA-paid community all-cause hospital admissions or ED visits was similar prior to service initiation and lower after service initiation, compared to enrollees in other VHA-paid personal care services programs (Figure 1A and 3A). For urban enrollees, the incidence of VHA or VHA-paid community all-cause hospital admissions and ED visits was similar across the VDC and comparison groups before and after service initiation (Figure 1B and 3B). For VDC enrollees residing in rural and urban areas, the incidence of VHA or VHA-paid community nursing home admissions was higher than comparison individuals before service initiation and similar to the incidence in the comparison groups after service initiation (Figure 2A and 2B).

Table 1.

Patient Characteristics and Standardized Differences

Variable	Rural/Urban Location	Comparison group at inactive sites	Comparison group at active sites	VDC group	Std. Diff. (%)
Gender: Male (%)	Rural	95.8	95.4	96.4	−4.9
	Urban	93.3	93.8	93.4	1.6
	Total	94.3	94.3	94.6	−1.1
Race: White (%)	Rural	97.5	98.0	98.8	−6.5
	Urban	97.4	96.6	98.1	−9.3
	Total	97.5	97.1	98.4	−9.0
Enrollment Priority 1a (%)	Rural	15.6	12.2	19.1	−19.2
	Urban	14.6	15.5	20.0	−11.9
	Total	15.0	14.3	19.6	−14.1
Medicaid Eligible (%)	Rural	0.8	0.9	0.9	0.1
	Urban	0.6	0.8	0.6	1.8
	Total	0.7	0.8	0.7	0.9
Aid and Attendance (%)	Rural	12.6	11.8	16.8	−14.2
	Urban	14.4	15.8	21.7	−15.2
	Total	13.7	14.5	19.8	−14.1
TBI (%)	Rural	1.5	1.6	2.1	−3.5
	Urban	1.5	1.7	2.8	−7.6
	Total	1.5	1.6	2.5	−6.0
Dementia (%)	Rural	24.2	24.2	23.3	2.1
	Urban	27.7	28.1	29.0	−2.0
	Total	26.3	26.8	26.7	0.1
SCI (%)	Rural	2.2	2.1	6.7	−22.4
	Urban	2.5	3.8	7.0	−14.0
	Total	2.4	3.2	6.9	−16.6
Age (Mean, SD)	Rural	76.6 (11.7)	77.3 (11.6)	76.1 (12.7)	9.7
	Urban	76.2 (12.3)	77.0 (12.5)	72.4 (14.9)	33.8
	Total	76.3 (12.1)	77.1 (12.2)	73.9 (14.2)	24.5
CAN Score (Mean, SD)	Rural	0.1 (0.2)	0.1 (0.2)	0.1 (0.2)	−2.0
	Urban	0.1 (0.2)	0.1 (0.1)	0.1 (0.1)	8.9
	Total	0.1 (0.2)	0.1 (0.1)	0.1 (0.1)	4.2
NOSOS Risk Score (Mean, SD)	Rural	3.2 (3.1)	2.9 (2.9)	3.5 (3.6)	−16.6
	Urban	3.3 (3.1)	3.1 (3.2)	4.0 (4.4)	−24.4
	Total	3.3 (3.1)	3.0 (3.1)	3.8 (4.1)	−21.1
Elixhauser Comorbidity Score (Mean, SD)	Rural	4.6 (3.0)	4.2 (3.0)	4.3 (3.0)	−1.8
	Urban	4.5 (3.0)	4.1 (3.1)	4.3 (3.2)	−3.8
	Total	4.5 (3.0)	4.2 (3.0)	4.3 (3.1)	−3.2
Mortality in the 12 months follow-up period (%)	Rural	22.5	21.6	24.0	−5.8
	Urban	21.9	21.8	21.7	0.1
	Total	22.2	21.7	22.7	−2.3

NOTE: The standardized differences are calculated between the comparison group at active sites and the Veteran-Directed Care (VDC) group, for Veterans living in the rural area, in the urban area and both. Comparison group at inactive sites: rural: N = 6,043; urban: N = 9,278; total: N = 15,321; comparison group at active sites: rural: N = 7,217; urban: N = 13,893; total: N = 21,110; VDC group: rural: N = 398; urban: N = 566; total: N = 964. TBI: Traumatic Brain Injury; SCI: Spinal Cord Injury; CAN Score: Care Assessment Need Score; NOSOS Risk Score: Risk adjustment for VHA patients.

Figure 1.

A Percent of rural Veterans with acute hospital admission(s) in each month

B Percent of urban Veterans with acute hospital admission(s) in each month

Figure 3.

A Percent of rural Veterans with Emergency Room visit(s) in each month

B Percent of urban Veterans with Emergency Room visit(s) in each month

Figure 2.

A Percent of rural Veterans with nursing home admission(s) in each month

B Percent of urban Veterans with nursing home admission(s) in each month

In our primary model specifications, both urban and rural VDC recipients were significantly less likely to be admitted to VHA or VHA-paid community nursing homes, compared to those receiving other VHA-paid personal care services (rural: incremental effect = −0.22, [ −0.30, −0.14]; urban: incremental effect = −0.14, [−0.20, −0.07]; Table 2). Rural, but not urban, VDC enrollees had significantly fewer all-cause VHA or VHA-paid community hospital admissions and VHA or VHA-paid community emergency department visits over time, relative to recipients of other VHA-paid personal care services (acute care, rural: incremental effect = −0.07, 95% CI = [−0.14, −0.01], urban: incremental effect = −0.01, [−0.06, 0.03]; ED visits, rural: incremental effect = −0.08, [−0.14, −0.02], urban: incremental effect = 0.01, [−0.03, 0.05]; Table 2). Odds ratios for these effects are available in Supplementary Table S1.

Table 2.

Relationships among VDC receipt and changes in healthcare use over time

Model		All-Cause Hospital Admissions
	Rural			Urban
	Dy/dx	95% Confidence Interval	p-value	Dy/dx	95% Confidence Interval	p-value
Unmatched fixed effects	−0.07	−0.13, −0.01	0.026	−0.01	−0.06, 0.03	0.510
Matched fixed effects	−0.08	−0.13, −0.02	0.012	−0.02	−0.06, 0.02	0.318
Model		Nursing Home Admissions
	Rural			Urban
	Dy/dx	95% Confidence Interval	p-value	Dy/dx	95% Confidence Interval	p-value
Unmatched fixed effects	−0.22	−0.30, −0.14	< 0.001	−0.14	−0.20, −0.07	< 0.001
Matched fixed effects	−0.19	−0.27, −0.11	< 0.001	−0.11	−0.18, −0.05	0.001
Model		ER Visits
	Rural			Urban
	Dy/dx	95% Confidence Interval	p-value	Dy/dx	95% Confidence Interval	p-value
Unmatched fixed effects	−0.08	−0.14, −0.02	0.008	0.01	−0.03, 0.05	0.534
Matched fixed effects	−0.08	−0.14, −0.02	0.009	0.02	−0.02, 0.06	0.433

The incremental effects from the sensitivity analysis modeling the interaction term between group and post were consistent with the results in the main analysis. (all-cause hospital admissions, rural: incremental effect = −0.07, 95% CI = [−0.13, −0.01], urban: incremental effect = −0.01, [−0.06, 0.03]; nursing home admissions, rural: incremental effect = −0.21, [ −0.28, −0.14]; urban: incremental effect = −0.13, [−0.20, −0.06]; ED visits, rural: incremental effect = −0.08, [−0.14, −0.02], urban: incremental effect = 0.01, [−0.03, 0.05])

We repeated our analyses for VHA acute care admissions, nursing home admissions and ED visits over matched cohorts of different VHA-paid personal care service recipients; including hospice utilization as an additional outcome; adopting a different definition of the outcome period; excluding VDC enrollees without any past VHA-paid personal care services; excluding decedents; excluding Veterans receiving services in less than 3 or 6 months in the outcome period; excluding Veterans with SCI. As anticipated, for rural and urban enrollees, the receipt of VDC was not associated with differences in hospice utilization, compared to the receipt of other VHA-paid care services. The results of other sensitivity analyses were consistent with the results in the main analysis (Supplementary Table S1).

Discussion

For rural and urban recipients, we estimated how trajectories in VHA-paid health care use from before to after enrollment in VDC, a self-directed services program, compared to trajectories in health care use among enrollees in other VHA-paid home- and community-based personal care services programs. Among a large cohort of Veterans receiving VHA-paid personal care services, we found that both rural and urban VDC recipients had fewer VHA or VHA-paid community nursing home admissions over time, compared to recipients of other VHA-paid personal care services. In addition, we found that rural, but not urban, VDC enrollees had fewer VHA or VHA-paid community acute care admissions and VHA or VHA-paid community emergency department visits over time.

To our knowledge, this is the first study to demonstrate that VDC recipients are less likely than recipients of other VHA-paid personal care services to have a VHA or VHA-paid community nursing home admission. Benefits were observed regardless of the urban or rural location and corresponded to the goal of the VDC program¹⁸ and previous findings that self-directed service programs reduced unmet needs for long-term care services and were related to less nursing home use^12–14. Future analyses should explore the degree to which nursing home admissions outside of the VHA change among VDC recipients, but the implementation of the VDC program has the potential to help older Veterans who are at risk of nursing home placement to remain in the community.

We observed fewer VHA or VHA-paid community acute care admissions and emergency department visits over time among rural VDC recipients relative to rural recipients of other VHA-paid personal care services. With a focus on rural/urban variation of the changes in health care use by recipients of VDC and other VHA-paid personal care programs, our results expand on the findings of a previous analysis, which reported that the VDC program did not have an overall effect on changes in the incidence of acute care admissions, relative to other VHA-paid personal care services programs.²⁰ Our acute care and emergency department visits findings are consistent with previous findings from the Cash and Counseling (C&C) Evaluation which observed an effect of the C&C program in Arkansas on health care utilization while the C&C programs in other states did not have an overall effect. The greater benefit observed in Arkansas corresponded to the fact that almost half of the participants lived in rural areas.¹²

The VDC program has unique attributes (e.g., allows the hiring of family or neighbors as caregivers and provides medical equipment), which might overcome the challenges in long-term care access that older adults living in rural areas face.^15,16 Our findings reveal that the flexibility and choice in care options provided by the VDC program may uniquely benefit enrollees living in rural areas.

This study has several limitations. First, we did not include non-VA paid care (hospital admission, nursing home admission, and ED visits) in the definition of the outcomes. In future work, we will examine the impact of the VDC program on healthcare use both in and outside the VHA. Second, our study does not compare the costs of administering the VDC program and other VHA-paid personal care services programs because enrollee-specific spending data, which depend on the patient’s case-mix level and the hours of care that the patient receives per week^26–27, are not readily available. Future work should examine costs associated with operating VHA-paid personal care services programs as well as costs associated with health care use. In addition, the sample size of the VDC group was relatively small, precluding more precise coding of rurality. However, to our knowledge, this is the first examination of the relative health care use patterns among VDC and other VHA-paid personal care service recipients in rural and urban areas.

Furthermore, the degree to which our findings are due to the VDC program or due to the presence of a strong family caregiver is unknown. Older adults at risk of nursing home placement often receive care from unpaid family caregivers.²⁸ The presence of a family caregiver is associated with decreased health care utilization, improved quality of life, and a lower risk of adverse health outcomes.^29–30 Although other VHA-paid care programs also involve family caregivers, the VDC program has a design that supports enrollees and their family caregivers through monthly budgets and long-term services and supports options counseling, so it is possible that VDC enrollees have family caregivers who are typically more involved in their care, regardless of participation in the program. Our study is unable to distinguish between the effects of the VDC program and the effects of possible between-program differences in the involvement of a family caregiver on the enrollee’s health care use. However, the fact that other VHA-paid personal care services programs also often involve family caregivers reduces the chance that the effects we observe are solely due to the presence of a family caregiver.

Our findings suggest that the VDC program is an appropriate and beneficial care option for Veterans with multiple chronic conditions and/or cognitive impairment and may be particularly beneficial for Veterans living in rural areas. Geographical variation in the effects of VDC on hospital and ED utilization indicates heterogeneity in the effects of the VDC program. VDC resources may be best targeted to Veterans living in rural areas who are uniquely likely to benefit from the VDC program. Given the differences observed in rural and urban health and healthcare utilization, our findings suggest that self-directed programs may alleviate some of the rural/urban disparities observed in prior literature.^1–3

Conclusion

VDC recipients had fewer incidents of potentially avoidable VHA or VHA-paid community health care use, compared to similar Veterans receiving other VHA-paid personal care services. The VDC program was associated with reduced VHA or VHA-paid community nursing home admissions regardless of the urban or rural location, compared to other VHA-paid personal care services. These benefits were more pronounced among rural VDC recipients than urban VDC recipients.

Supplementary Material

Appendix

Supplementary Table S1 VDC Sensitivity Analysis Output Summary

Key Points:

• Implementation of the VDC program has the potential to reduce VHA-paid nursing home placement.

• Variations in trajectories in health care use exist between rural and urban recipients of the VDC and other VHA-paid personal care services programs.

• The flexibility and care options provided by the VDC program may uniquely benefit enrollees living in rural areas.

Why does this paper matter:

This study reveals that the VDC program reduced VHA-paid nursing home admissions regardless of rural or urban location, compared to other VHA-paid personal care services programs. The benefits were more pronounced among rural recipients than urban recipients of the VDC program.