Abstract
We examined data for all veterans who completed the Veterans Health Administration’s national homelessness screening instrument between October 1, 2012, and January 10, 2013. Among veterans who were not engaged with the US Department of Veterans Affairs homeless system and presented for primary care services, the prevalence of recent housing instability or homelessness was 0.9% and homelessness risk was 1.2%. Future research will refine outreach strategies, targeting of prevention resources, and development of novel interventions.
Homelessness prevention is an essential component of the US Department of Veterans Affairs’ (VA’s) strategy to end homelessness among veterans by the end of 2015. In support of this strategy, the VA has unveiled Supportive Services for Veteran Families, providing more than $500 million to support local grantees’ provision of supportive services and financial assistance to low-income veterans living in or transitioning to permanent housing. As has been addressed elsewhere, efficient and effective targeting of homelessness prevention resources is challenging for multiple administrative and social reasons.1–3 In an effort to improve the ability to identify veterans who are at risk for homelessness—or experiencing homelessness but not accessing services—the National Center on Homelessness Among Veterans developed the Homelessness Screening Clinical Reminder for use in patients’ electronic medical records to conduct a universal screen for housing instability and risk (A. E. Montgomery, PhD, J. D. Fargo, PhD, V. R. Kane, MSS, and D. P. Culhane, PhD, unpublished data, 2013).
The objective of this national, health system-based screening instrument is to enhance the rapid identification of veterans and their families who have very recently become homeless or are at imminent risk of homelessness and ensure that they are referred for the appropriate assistance. This study is the first examination of national screening data for homelessness and homelessness risk within the Veterans Health Administration (VHA) and provides an overview of the prevalence of homelessness risk among veterans seeking health-related services through VHA.
METHODS
The Homelessness Screening Clinical Reminder assesses veterans’ current experience of housing instability (“In the past 2 months, have you been living in stable housing that you own, rent, or stay in as part of a household?”) as well as veterans’ imminent risk of homelessness (“Are you worried or concerned that in the next 2 months you may NOT have stable housing that you own, rent, or stay in as part of a household?”). Veterans excluded from screening are those who received homeless services through VA Specialized Homeless Programs within the previous 6 months or were receiving long-term or palliative care. We extracted data from the VA Corporate Data Warehouse for all veterans who completed the Homelessness Screening Clinical Reminder between October 1, 2012, and January 10, 2013, and analyzed these data using SAS version 9.3 (SAS Institute, Cary, NC) and R version 2.15 (Statistics Department, University of Auckland, New Zealand).
RESULTS
A total of 1 422 038 veterans were presented with the screener. Of these, we excluded data for 1342 because of coding errors or incomplete screens, and 15 were excluded for being out of the age range (i.e., < 18 or > 115 years), resulting in the loss of 0.1% of the sample. Of the remaining participants, a total of 1 398 925 veterans completed the screen and 21 756 veterans declined the screen because they were already receiving housing assistance (2802; 12.9%), living in a nursing home (3904; 17.9%), or unable or unwilling to respond (15 050; 69.2%). The screened sample was 97% male with a mean age of 62.6 years (SD = 15.0).
Of those screened, 12 754 reported current unstable housing or homelessness (0.9%) and 17 211 (1.2%) reported being at risk for housing instability, resulting in a total of 29 965 positive screens (2.1%) and 1 368 302 negative screens (97.9%; Table 1.) Female veterans reported significantly higher levels of both housing instability (1.1% vs 0.9%) and homelessness risk (1.9% vs 1.2%; χ21 = 30.58; P < .001). Veterans younger than 64 years made up the majority of positive screens; veterans between the ages of 35 and 54 years accounted for 36.7% of positive screens (χ24 = 9662.49; P < .001).
TABLE 1—
Veteran Characteristics, Screening Disposition, and Living Situation: Philadelphia Veterans Affairs Medical Center, PA; October 1, 2012–January 10, 2013
| Positive Screen, No. (%) |
|||
| Characteristic | Housing Instability or Homeless | Homelessness Risk | Negative Screen, No. (%) |
| Total | 12 754 (0.9) | 17 211 (1.2) | 1 368 302 (97.9) |
| Gender | |||
| Female | 1213 (1.1) | 1980 (1.9) | 104 127 (97.0) |
| Male | 11 541 (0.9) | 15 231 (1.2) | 1 216 174 (97.9) |
| Age, y | |||
| 18–34 | 1663 (1.8) | 2050 (2.2) | 87 835 (95.9) |
| 35–54 | 4660 (1.9) | 6326 (2.5) | 240 714 (95.6) |
| 55–64 | 4467 (1.2) | 6075 (1.6) | 371 771 (97.2) |
| 65–74 | 1452 (0.4) | 2093 (0.6) | 372 804 (99.1) |
| ≥ 75 | 510 (0.2) | 664 (0.2) | 295 143 (99.6) |
| Living situationa | |||
| House—with subsidy | 325 (2.5) | 1092 (6.3) | NA |
| House—no subsidy | 2704 (21.2) | 10 408 (60.5) | NA |
| With friend or family | 4719 (37.0) | 3612 (21.0) | NA |
| Motel or hotel | 671 (5.3) | 155 (0.9) | NA |
| Institution | 484 (3.8) | 127 (0.7) | NA |
| Shelter | 737 (5.8) | 58 (0.3) | NA |
| Street | 1321 (10.4) | 80 (0.5) | NA |
| Other situation | 1649 (12.9) | 1679 (9.8) | NA |
Note. NA = not applicable. Row and column totals may not equal 100% because of missing data. The sample size was n = 1 398 925.
Percentages for living situation are based on screening disposition (i.e., column percentages); living situation is only available for veterans who screened positive. Note that veterans are asked to report their living situation for most of the 2-mo period before date of screening; therefore, some contradiction may exist between their screening disposition and living situation (e.g., some veterans who reported housing instability also reported living in their own unsubsidized housing for the majority of the 2-month period).
Table 1 compares respondents’ current living situation (i.e., where the veteran lived for most of the previous 2 months) for veterans who screened positive. Both groups were highly likely to be living in unsubsidized housing (21.2% of those who were homeless, 60.5% of those at risk) or in a doubled-up situation with friends and family (37.0% of those who were homeless, 21.0% of those at risk). Veterans who screened positive for housing instability were less likely to be living in housing, and those who screened positive for risk were more likely to be in housing (χ27 = 6749.98; P < .001).
DISCUSSION
Initial data from the first quarter of the VHA national-level screener for homelessness and risk for homelessness showed prevalence rates of 0.9% and 1.2%, respectively, among veterans who have not been recently involved with the VA homeless system and were presenting for VA primary care services. Not only do these initial data support the estimation of the systemwide prevalence of homelessness and homelessness risk, they also help identify areas for future research to inform outreach, targeting, and homelessness prevention interventions. Specifically, future work will focus on sociodemographic characteristics such as race, ethnicity, gender, and medical and behavioral health diagnoses and the implications for effective targeting of homelessness prevention resources.
Acknowledgments
This project was funded by the US Department of Veterans Affairs, National Center on Homelessness Among Veterans.
Christopher B. Roberts provided assistance with data management and analytic support.
Note. The contents of this article do not necessarily represent the views of the US Department of Veterans Affairs or the US Government.
Human Participant Protection
Institutional review board (IRB) approval was obtained from the Philadelphia Veterans Affairs Medical Center IRB for secondary analysis of existing administrative data.
References
- 1.Burt MR, Pearson CL, Montgomery AE. Community-wide strategies for preventing homelessness: recent evidence. J Prim Prev. 2007;28(3-4):213–228. doi: 10.1007/s10935-007-0094-8. [DOI] [PubMed] [Google Scholar]
- 2.Lindblom EN. Toward a comprehensive homelessness prevention strategy. Housing Policy Debate. 1991;2(3):957–1025. [Google Scholar]
- 3.Shinn M, Baumohl J, Hopper K. The prevention of homelessness revisited. Analyses Soc Issues Public Policy. 1991:95–127. [Google Scholar]