Change in Outpatient Care Following Migration Among Veterans with Experience of Housing Instability

Individuals who experience housing instability are more likely to have adverse health outcomes, particularly in rural areas where services and housing options are often scarce.¹ Veterans—who are both more likely to experience housing instability² and reside in rural areas³ compared with non-Veterans—may face unique challenges accessing health services.⁴ Although the US Department of Veterans Affairs (VA) offers a nationwide network of services in rural and urban settings, Veterans who experience housing instability and migrate (i.e., residential relocation of > 40 miles or a change in rural/urban residence) to or within a rural area have lower use of homeless services, outpatient care, emergency department visits, and inpatient admissions compared to those who migrate to an urban area.⁴ Understanding how patterns of services utilization change as Veterans relocate between rural and urban areas is crucial to ensure that Veterans with experience of housing instability access necessary care (Table 1).

Table 1.

Change in Outpatient Healthcare Utilization and Adjusted Difference-In-Difference Estimates, Pre-/Post-Migration to a Rural vs. Urban Destination (N = 81,620)

Pre-Migration Residence/Facility	Post-Migration Residence	Adjusted Population Means for Primary Care Utilization			Difference-in-Difference Estimate (95% CI)	p-value
		Pre-Migration	Post-Migration	Δ
Rural Residence/Rural Facility	Rural	3.40	3.33	− 0.07	− 0.45 (− 0.67, − 0.24)	<.001
	Urban	3.23	3.61	0.38
Rural Residence/Urban Facility	Rural	2.39	2.63	0.23	− 0.22 (− 0.35, − 0.09)	.001
	Urban	2.45	2.90	0.45
Urban Residence/Rural Facility	Rural	3.37	3.40	0.03	− 0.61 (− 0.97, − 0.25)	<.001
	Urban	3.04	3.68	0.64
Urban Residence/Urban Facility	Rural	2.81	2.91	0.10	− 0.39 (− 0.45, − 0.32)	<.001
	Urban	2.81	3.30	0.49
Pre-Migration Residence/Facility	Post-Migration Residence	Adjusted Population Means for Mental Health Care Utilization			Difference-in-Difference Estimate (95% CI)	p-value
		Pre-Migration	Post-Migration	Δ
Rural Residence/Rural Facility	Rural	12.02	10.99	− 1.04	− 1.54 (− 2.44, − 0.64)	<.001
	Urban	11.73	12.24	0.50
Rural Residence/Urban Facility	Rural	11.61	10.86	− 0.75	− 1.07 (− 1.6, − 0.52)	<.001
	Urban	12.55	12.87	0.31
Urban Residence/Rural Facility	Rural	13.95	10.87	− 3.08	− 2.35 (− 4.1, − 0.55)	.010
	Urban	12.84	12.11	− 0.72
Urban Residence/Urban Facility	Rural	11.89	10.82	− 1.06	− 1.20 (− 1.46, − 0.93)	<.001
	Urban	12.37	12.51	0.14
Pre-Migration Residence/Facility	Post-Migration Residence	Adjusted Population Means for Specialty Outpatient Care Utilization			Difference-in-Difference Estimate (95% CI)	p-value
		Pre-Migration	Post-Migration	Δ
Rural Residence/Rural Facility	Rural	5.46	5.52	0.07	− 0.29 (− 0.82, 0.23)	.270
	Urban	5.64	5.99	0.36
Rural Residence/Urban Facility	Rural	4.83	4.67	− 0.16	− 0.71 (− 1.02, − 0.41)	<.001
	Urban	4.80	5.34	0.55
Urban Residence/Rural Facility	Rural	6.99	5.84	− 1.14	− 1.57 (− 2.53, − 0.62)	.001
	Urban	6.00	6.44	0.43
Urban Residence/Urban Facility	Rural	5.73	5.34	− 0.39	− 0.68 (− 0.83, − 0.54)	<.001
	Urban	5.79	6.09	0.30
Pre-Migration Residence/Facility	Post-Migration Residence	Adjusted Population Means for Ancillary Outpatient Care Utilization			Difference-in-Difference Estimate (95% CI)	p-value
		Pre-Migration	Post-Migration	Δ
Rural Residence/Rural Facility	Rural	5.95	5.99	.05	− 0.83 (− 1.20, − 0.46)	<.001
	Urban	5.66	6.54	0.88
Rural Residence/Urban Facility	Rural	5.24	5.44	0.21	− 0.51 (− 0.74, − 0.28)	<.001
	Urban	5.41	6.12	0.72
Urban Residence/Rural Facility	Rural	6.97	6.91	-.06	− 1.01 (− 1.59, − 0.42)	<.001
	Urban	6.29	7.23	0.95
Urban Residence/Urban Facility	Rural	5.74	5.71	− 0.03	− 0.75 (− 0.85, − 0.64)	<.001
	Urban	5.69	6.41	0.72

Facility indicates where the Veteran was identified as experiencing housing instability; migration is based on change in residential location. Bold indicates p <.05. Analyses control for number of quarters between homelessness and migration, sex; age; race; ethnicity; marital status; history of combat exposure and military sexual trauma; service connection; mental health, medical, and substance use comorbidities; and prior health care use

The present study extends a prior analysis⁴ of services use among Veterans with experience of housing instability who migrate by examining changes in four types of outpatient care, offering insight into the complex association among mobility, homelessness, and healthcare access for this vulnerable population.

METHODS

As in prior work, this study uses data from VA administrative sources for 81,620 Veterans with an indicator of incident homelessness at a rural or urban facility during 1/1/2014–12/31/2018 and who migrated.⁴ We compared changes in four types of outpatient service utilization—primary care, mental health, specialty outpatient care (i.e., surgery, rehabilitation), and ancillary care (i.e., diagnostic, other procedures)—from the two quarters pre-migration to the two quarters post-migration. We used a difference-in-differences (D-I-D) approach and conducted distribution-appropriate mixed effect repeated-measures linear regression models stratified by the index residence/facility variable and controlling for covariates (i.e., demographics; military history; mental health, medical, and substance use comorbidities; prior health care use). This study was approved by the Institutional Review Board at Birmingham, Alabama Veterans Affairs Health Care System.

RESULTS

Veterans who migrated to or within rural areas had decreased utilization of mental health and specialty outpatient care while utilization of primary and ancillary care remained approximately the same. Veterans who migrated to or within urban areas saw increased utilization across all care types. Results from the D-I-D analysis confirmed these descriptive findings: Veterans who migrated to or within rural areas had significantly lower utilization of each type of outpatient care assessed here compared with Veterans who migrated to or within urban areas. Veterans identified as experiencing housing instability at a rural facility while residing in a rural area who migrated to or within a rural area (vs. an urban area) had 0.45 fewer primary care visits, 1.54 fewer mental health outpatient visits, and 0.83 fewer ancillary care visits in the two quarters post-migration (p < 0.001). Specialty outpatient care utilization was also lower although this result was not statistically significant.

DISCUSSION

Housing instability among Veterans impacts their access to healthcare, with notable rural/urban disparities. The findings from this study underscore the association between geographic factors on healthcare accessibility⁵ and extend a prior study by focusing on specific types of healthcare (e.g., primary care, mental health). These findings are consistent with the research conducted by Miller et al., which indicated that while Veterans experiencing housing instability often seek increased mental health services, their overall outpatient care diminishes, particularly after relocating to rural areas.⁶ In the present study, there was a clinically significant decrease in the number of mental health visits following migration. This urban–rural divide necessitates interventions tailored to specific barriers faced by rural Veterans, combining policy reforms with the increased use of telehealth, which can reduce the impact of geographic barriers on access to care.^6,7

Limitations of this study include lack of data for Veterans with experience of housing instability who did not engage with VA care. Future research should focus on the effects of migration on healthcare utilization, the potential of integrated care models to address disparities in different geographic contexts, and the driving factors behind migration. Exploring the role and availability of community-based support systems and their effectiveness in rural settings could inform policies aimed at bridging service gaps. By focusing on these efforts, stakeholders can work toward creating a more equitable healthcare system and improving health outcomes for Veterans experiencing housing instability.

Funding

This study was funded by the U.S. Department of Veterans Affairs, Office of Rural Health, Veterans Rural Health Resource Center – White River Junction (OMAT 15986, PI Montgomery) and the Health Services Research & Development program (IIR- 19–355, M-PI Montgomery).

Declarations:

Conflict of Interest:

The authors declare that they do not have a conflict of interest.