Preventing Homelessness Through the National Call Center for Homeless Veterans: Analysis of Calls and Service Referrals

The NCCHV provided data for contacts made to the NCCHV and stored within the Microsoft Dynamics Customer Relationship Management software system between November 30, 2018, and October 31, 2020. When an individual contacts the NCCHV, responders are trained to engage the individual and ask a series of scripted questions to assess and collect data on background characteristics, housing status, risk for homelessness, and health and social needs; all data fields may be asked, but individuals who contact the NCCHV may choose not to provide certain information. All VA medical facilities have a designated VHA homeless program point-of-contact (POC) who responds to referrals from the NCCHV and provides follow-up contact and a plan to coordinate appropriate services for the veteran. Information about completed referrals is returned to the NCCHV electronically, indicating the outcome and resolutions of the case.

During the study period, the NCCHV received 266,100 telephone calls and chat messages (approximately 11,570 per month) across regions partnering with >160 U.S. VA medical facilities. Because most help-seeking veterans made calls, we combined calls and chat messages and refer to them here as “contacts.” As detailed in a diagram in an online supplement to this report, our main analysis focused on the samples of total unique veterans (N=110,197) and unique veterans who engaged with VHA health care (N=48,808).

We merged NCCHV data with VHA medical records for the year before and after veterans’ contact with the NCCHV during the study period. Data from VA medical records contained information on health care and homeless service use stored in the VA’s Corporate Data Warehouse and Homeless Operations Management Evaluation System and included individual veteran identifiers (i.e., names, telephone numbers, and Social Security numbers). Using these merged data, we examined whether veterans were connected to any VA services before and after their NCCHV contact. Because some veterans made multiple contacts with the NCCHV, we retained the date of their first contact during the study period as the index date. We limited these analyses to veterans who made their first NCCHV contact before April 1, 2020, to allow for a full year of observation after their contact (N=48,808). To examine changes in NCCHV contacts during the COVID-19 pandemic, we compared the total number of NCCHV contacts and number of unique veterans who contacted the NCCHV 6 months before the COVID-19 pandemic with those 6 months during the pandemic (with March 13, 2020, designated as the start of the pandemic in the United States). Given the large sample sizes involved in the aforementioned analyses, we did not rely only on statistical significance but also examined the change in percentages for meaningfulness of differences. This study was a quality improvement project initiated by the VHA Homeless Program Office in coordination with the NCCHV and was therefore deemed exempt from review by the institutional review board of South Texas VA Health Care System.

Among the total sample of 266,100 NCCHV contacts during the period studied (corresponding to approximately >50,000 calls per year), 61.7% (N=164,203) were from veterans themselves, 5.9% (N=15,757) from VA staff members, 5.9% (N=15,593) from community providers, 3.0% (N=8,005) from veteran spouses, 1.9% (N=4,943) from friends, and the remaining from interested third parties, siblings, parents, and others. Regarding contacts’ locations, most (87.2%, N=231,970) were made from cell phones, 9.1% (N=24,285) by Web Chat visitors, 3.0% (N=8,108) from community or government agencies, and <1% (N=1,727) from some other location (e.g., hotel, pay phone, or shelter). The total number of NCCHV contacts in the 6 months before the COVID-19 pandemic was 38,147 (involving 26,875 unique individuals), and the total number of NCCHV contacts in the first 6 months of the COVID-19 pandemic was 35,085 (24,575 unique individuals).

Among 266,100 NCCHV contacts, 110,197 involved unique veterans. On the basis of these unique veteran contacts and the NCCHV responders’ assessment of the veterans’ current housing situations, 20.1% (N=22,189) were recorded as homeless, 69.6% (N=76,732) as at risk for homelessness, 5.1% (N=5,669) as not homeless or at risk, 2.2% (N=2,370) as not homeless or risk unknown, and 3.5% (N=3,819) were missing a designation. Among veterans who made multiple contacts to the NCCHV, the mean length of time between calls was 81 days. Half (50.0%, N=55,107) of the veterans were ages 30–59 years, and only 9.6% (N=10,571) were ages <30 years, and 15.4% (N=17,017) were >70 years. Most veterans identified as male, with 13.1% (N=14,461) identifying as female (see first table of the online supplement). Many veterans reported receiving public support income (e.g., Social Security, Supplemental Security Insurance [SSI], or VA disability and pension), with 35.9% (N=39,577) reporting SSI and 13.7% (N=15,122) reporting VA disability or VA pension. In addition, veterans reported several legal concerns, most frequently eviction (7.6% N=8,381) and child support issues (4.7%, N=5,147).

As shown in the second table in the online supplement, nearly all contacts from unique veterans (90.2%, N=99,371) resulted in a referral and in the veteran being transferred to a local homeless POC. The second most common outcome, for about 7.1% (N=7,847) of veterans, was responding to counseling during the contact and not requiring further intervention. All other outcomes were very rare, including disconnected contacts, contacts refusing referrals, or transfers to the suicide hotline.

Examination of VHA medical records revealed that of the 110,197 unique veterans who contacted the NCCHV, 65,587 (59.5%) had a VHA medical record. The age and gender compositions of this subsample were similar to those of the overall sample of veterans who contacted the NCCHV. Additional data gleaned from VHA medical records revealed that the race-ethnicity breakdown of the subsample of VHA users was 41.5% (N=27,241) non-Hispanic White, 39.2% (N=25,691) non-Hispanic Black, 8.1% (N=5,305) Hispanic, and 4.0% (N=2,648) mixed or other race.

Table 1 compares the use of VHA homeless and health care services among veterans during the year before and after their initial contact with the NCCHV (N=48,808 with a full year of follow-up data). Of veterans who contacted the NCCHV and had a VHA medical record, 25.9% had used a VHA homeless program the year before their NCCHV contact, a proportion that increased to 81.3% within a year post-NCCHV contact. We noted major increases in the use of all individual VHA Homeless Programs after an NCCHV contact. In addition, 78.4% of veterans who contacted the NCCHV and were VHA users accessed VHA health care services during the year before their NCCHV contact, a proportion that increased to 93.5% within the year post-NCCHV contact. Across individual health care services, the greatest increases were in outpatient medical care (76.8%–93.1%) and mental health care (43.3%–51.4%).

In this first study of the NCCHV, we found that the NCCHV received on average >50,000 contacts from veterans annually seeking assistance and services to address housing instability and homelessness. We did not detect an increase in NCCHV contacts during the first 6 months of the COVID-19 pandemic. Notably, most contacts to the NCCHV were made by veterans experiencing an increased risk for homelessness, indicating the importance of the NCCHV as a resource for homelessness prevention. As described previously (6–9), preventing homelessness among veterans is a policy area of particular importance to the VA, and services offered through the NCCHV appear to be, at least in part, addressing this need.

Many contacts to the NCCHV resulted in a referral to services, although 40.5% (N=44,610) did not have a linked VHA medical record 1 year after their first NCCHV contact, so we could not assess their receipt of VHA services. Several reasons may explain why veterans did not use VHA services after a referral. For example, a veteran’s issues may have been resolved during the NCCHV contact, and no follow-up was needed from the veteran’s perspective, a scenario that is somewhat consistent with research on a national suicide prevention lifeline that found that the most prevalent barrier to service use was the caller’s perception of the severity of the problem and belief that the problem could be handled with services (11). Moreover, some NCCHV veterans may have been connected to a non-VHA community provider and received services there rather than through a VHA provider. Finally, another possible reason may be that veterans did not want additional services or that some type of disconnect may have existed among the NCCHV referrals, VHA providers, and the veterans in making a referral to VHA services. We were not able to determine how much of a factor this third scenario might have been or what barriers to services might have played a role, a knowledge gap that deserve future study.

Of note, the results of our main analyses revealed that veterans who contacted the NCCHV and connected to VHA services significantly increased their use of VHA homeless programs within 1 year after their initial NCCHV contact. In addition, we observed small increases in their use of VHA inpatient mental health, substance use, and medical care services along with outpatient substance use services (<5% change) and moderate increases in the use of VHA outpatient and medical care services (>5% change). These findings are encouraging because they suggest that the NCCHV serves as an important “door” in a “no wrong door” approach for connecting veterans to needed services, both within and outside of the VHA. Although we had limited data on the use of specific services to resolve income and civil legal issues that veterans reported, the VHA directly offers or partners with community programs to offer financial assistance, supported employment, and medical-legal partnerships to assist with these issues (8, 12, 13), and veterans can be linked to these services through the NCCHV. Together, these findings underscore the array of VHA-based and community-partnered homeless, health care, and social services available to veterans who are experiencing homelessness or are at risk for homelessness.

Our study had several limitations. We did not have data on anonymous individuals who contacted the NCCHV, nor did we have data on the use of non-VHA homeless services that veterans may have used. We had only service use data on veterans enrolled in VA care, although it is important to study at-risk veterans who do not use the VA and the barriers to care they experience as well as institutional gaps in coordination or services. Despite these limitations, major strengths of our study were the use of national data from both the NCCHV and the VHA and 1-year follow-up after an NCCHV contact. This initial descriptive study points to many areas for further research, including studies with a longer follow-up period, data collection from non-VHA homeless services, and examination of effective communication and coordination methods provided by NCCHV responders.

Several lessons may be learned from this descriptive study of the NCCHV that may inform other systems of care. Many individuals at risk for homelessness contacted the NCCHV, even those who had used homeless programs before. Some issues could be rapidly resolved, providing support for some approaches that include homeless diversion and rapid resolution. Most NCCHV contacts were made from cell phones, suggesting the importance of access to technology. Given the importance of technology in modern society and concerns about a digital divide in rural or low-income regions (14), initiatives to provide technology, encourage its uptake, and improve technology literacy may have many positive downstream effects, including preventing homelessness.