Stakeholder perspectives on sustainment of Housing First in a VA permanent supportive housing program

Abstract

Objective

To evaluate the sustainment of Housing First (HF) implementation in a permanent supportive housing program for homeless‐experienced veterans, 5 years after practice implementation.

Study setting

From 2016 to 2017, primary data were collected from providers and veterans in the Department of Housing and Urban Development–VA Supportive Housing (HUD‐VASH) program at Los Angeles.

Study design

Guided by the integrated sustainability framework, we performed a mixed‐methods study to evaluate the sustainment of HF, an evidence‐based practice implemented to improve housing outcomes. To assess sustainment, we measured fidelity to HF in six of seven HUD‐VASH teams. These data were integrated with qualitative interviews with providers and veterans who described perceived sustainment to HF and contextual factors that supported or impeded sustainment.

Data collection

Fidelity to HF at 5 years after practice implementation, as a proxy for sustainment, was quantified via surveys with HUD‐VASH teams. HUD‐VASH providers (n = 51) and 31 veterans participated in semi‐structured interviews. Team‐based template analyses were used to develop an emergent understanding of stakeholder perspectives on HF sustainment.

Principal findings

Overall, HUD‐VASH teams reported HF sustainment. The lowest fidelity scores were found in the domains of client‐to‐staff ratios, frequency of client–provider contact, and time to housing. Qualitative findings indicated that outer contextual factors (e.g., housing scarcity) and organizational factors (e.g., staff turnover) impacted HF sustainment. Providers identified changes in leadership and unmet resource needs as impediments to practice sustainment. All stakeholders identified positively with the HF practice and believed that the approach benefited veterans.

Conclusions

This snapshot of HF sustainment demonstrates that this practice can be sustained over time. However, strong leadership, organizational resources, and community partnerships are needed. Adaptations to HF in response to outer contextual factors and organizational capacity may result in practice sustainment while allowing for flexibility in service provision.

What is known on this topic

• Housing First, an evidence‐based practice, was implemented at the VA in the early 2010s in its permanent supportive housing program, known as the Department of Housing and Urban Development–VA Supportive Housing (HUD‐VASH) program.

• Significant variability in Housing First fidelity has been documented in the literature.

• Little has been published about the sustainment of the Housing First practice after the withdrawal of intensive efforts to implement it in permanent supportive housing programs.

What this study adds

• This study measures Housing First fidelity 5 years after its implementation, which we conceptualize as a measure of sustainment, and assesses stakeholder perspectives on sustainment.

• It describes factors that can contribute to Housing First sustainment despite contextual barriers to sustainment, for example, a competitive, expensive rental market.

1. INTRODUCTION

Housing First (HF) is an evidence‐based practice that prioritizes rapid, permanent supportive housing for homeless persons by providing subsidies for independent housing, along with case management and linkages to services. ¹ A key component of HF is the absence of mandates for sobriety or mental health treatment, in contrast to traditional approaches to homelessness, which operationalize housing readiness via treatment compliance and/or sobriety. ² Additional key components of HF include the delivery of voluntary, community‐based, and recovery‐oriented services at a pace and intensity determined by clients, ² and the prioritization of housing for chronically homeless persons, including those with high psychosocial or biomedical vulnerabilities. ³ , ⁴ , ⁵ Table 1 details the key components of HF, which are often used to measure fidelity to this practice. A substantial body of research shows that HF increases access to housing, ⁶ , ⁷ , ⁸ reduces the use of costly emergency services, ⁹ , ¹⁰ , ¹¹ heightens perceived autonomy, and improves housing outcomes. ¹² , ¹³ As such, many housing programs for homeless persons have made significant efforts to implement HF; however, after these intensive implementation efforts cease, little is known about the sustainment of HF in these settings. ¹⁴

TABLE 1.

Key components of Housing First, used to assess fidelity to Housing First practice

Domain	Key components
Housing process and structure	Residents can choose their neighborhood, building, unit, furnishings, and decorations. Program provides housing subsidies or subsidized housing units. Rapid turnover from enrollment to permanent supportive housing (in less than a month). Residents live alone in scattered‐site permanent supportive housing, rather than in group‐setting and/or emergency, short‐term, or transitional housing. Residents pay 30% or less of their income towards rent.
Separation of housing and services	Housing enrollment and maintenance are not contingent on service use or symptom stability, solely meeting responsibilities of a standard lease. Residents have a choice in services, including pharmacotherapy and substance use disorder treatment. When residents lose their housing, the program continues to provide services and offers them a new unit, without requiring certain criteria to retain these services and housing.
Service philosophy	Residents have the right to choose, modify, or refuse services, including psychiatric or substance use disorder treatment. Residents are not required to abstain from alcohol or drugs, and providers support harm‐reduction approaches to minimize risk from substance use. Program uses motivational interventions to build rapport with participants who are difficult to engage in services. A recovery‐oriented treatment plan is co‐created with a resident and staff, and the program employs data to adjust services accordingly.
Service array	Program offers an array of services, including property management, rental payment assistance, cosigning of leases, substance use disorder treatment, community‐based employment, supported educational opportunities, volunteer services, coordinated medical services, and social integration assistance. A health care provider completes a monthly review of residents' symptoms and responses to medication, when applicable. Permanent supportive housing staff include a paid, certified peer support specialist or more for every 100 residents. Staff have received training in a strengths‐based approach and practice motivational interviewing in all aspects of treatment. The program consults on residents' hospital admission. and discharge planning, in cases of inpatient stays.
Team structure	The staff‐to‐resident ratio is 1–20 or fewer residents. Health care provider‐to‐patient ratio is 1–75 or fewer residents. Most residents have in‐person contact with staff at least once a month, preferably at residents' homes. Program staff meet at least 4 days per week. Team meetings are used for multiple functions, such as reviewing recent contact with residents, supporting residents' long‐term goals, developing staff schedules based on residents' needs, and so forth. Program includes residents on governing bodies, employs persons with lived experience in regular staff positions, offers regular opportunities for resident feedback, employs peer specialists, and has a formal grievance process for participants to express concerns or dissatisfaction.

Following a 2009 federal priority to end veteran homelessness, the Department of Housing and Urban Development (HUD) in collaboration with the Department of Veteran Affairs permanent supportive housing (VASH) program, hereafter known as HUD‐VASH, began robust efforts to implement HF. ¹⁵ , ¹⁶ HUD‐VASH was originally developed in the early 1990s to provide subsidized permanent housing and supportive services for homeless‐experienced veterans. ¹⁷ Historically, many elements of HUD‐VASH did not align with the HF model: for example, there was variability in treatment mandates for veterans who engaged in the program and differing practices with regard to prioritization of veterans who received services. ¹

Diverse strategies were used to implement HF in HUD‐VASH, including coalition building, training and technical assistance, site visits, and the use of expert consultants. ¹⁸ These concerted efforts to implement HF were temporally correlated with a dramatic increase in HUD‐VASH services; by 2020, HUD‐VASH had provided supportive services to nearly 80,000 VA health care eligible veterans. ¹⁹ This initiative led to a 49% reduction of homeless‐experienced veterans, from approximately 73,000 veterans counted on a single night in 2009 as compared with 36,000 veterans in 2020. ²⁰ At present, HUD‐VASH serves 7500 veterans in metropolitan Los Angeles. ²¹ HUD‐VASH teams, composed of social workers, nurses, and peer support specialists, provide supportive services in the field, including eligibility assessments, home visits, care coordination, and transportation to services. ⁵

As a part of HF's widespread implementation within and outside VA, practice adaptations have been made to fit varying settings and contexts. ²² , ²³ , ²⁴ , ²⁵ As intensive HF implementation efforts have largely ceased at VA Greater Los Angeles (VAGLA), additional research is needed to characterize practice sustainment and factors associated with sustainment. ¹⁴ , ²⁵ A salient challenge to measuring sustainment is limited resources for longitudinal assessments of fidelity and adaptation to organizational context. ¹⁴ Implementation fidelity, measured after intensive implementation efforts have ceased as a proxy for HF sustainment, seeks to assess the real‐world application of a practice as compared with the original implementation plan, as well as context‐specific practice adaptations that may evolve over time. ¹⁴ , ²⁵ , ²⁶ , ²⁷

This study was conducted 5 years after HF implementation at VAGLA; intensive implementation efforts, employing strategies including training and technical assistance, clinical supervision, and implementation consult, took place between 2016 and 2017. This study enabled a snapshot of HF years after these efforts were withdrawn, reflecting a practice adapted over time to regional issues, organizational structure, and community resources. Guided by the integrated sustainability framework, we used mixed methods to characterize HF sustainment, integrating an objective fidelity measure with provider and veterans' perspectives on HF. ²⁸ With the goal of enhancing HF sustainment longitudinally, this study seeks to link outer contextual factors to organizational issues that impact practice sustainment, identifying factors that support or impede HF sustainment as well as practice adaptations to the program's setting and context.

2. METHODS

2.1. Conceptual framework

This study is guided by the integrated sustainability framework, ²⁸ which depicts the relationships between outer contextual factors (e.g., sociopolitical context, external leadership, and priorities) and organizational factors, including inner contextual factors (e.g., leadership, organizational resources, and staffing/turnover), processes (e.g., partnership, training/supervision, practice evaluation, and adaptation), provider characteristics (e.g., skills and expertise), and intervention characteristics (e.g., perceived benefit/need and adaptability). ²⁸ These outer contextual factors and organizational factors contribute to practice sustainment, which here reflects continued HF implementation.

2.2. Study setting and design

This study was conducted in the HUD‐VASH program at VAGLA, which serves veterans in a vast metropolis known for its high cost of living, competitive rental markets, and housing scarcity. ²⁹ , ³⁰ We used a convergent design to analyze qualitative interviews with veterans and providers, along with self‐reported fidelity survey data from providers. All study procedures were approved by the VAGLA Institutional Review Board.

2.3. Qualitative interviews with veterans and providers

A licensed clinical social worker (S.H.) and research assistant (A.F.) recruited and interviewed veterans. Content analysis based on the integrated sustainability framework underpinned the study. Veterans (n = 31) were purposively sampled, recruited from a stratified random sample (n = 121) of participants from the VAGLA HUD‐VASH roster, with relatively equal distribution across Los Angeles's eight service planning areas, geographic regions designated by the county for health care planning purposes. No participant left the study after consenting. Interviews were conducted in person at VA offices. Veteran participant characteristics are described in Table 2; their demographics were representative of HUD‐VASH's client population. ²⁷ During their interviews, veterans receiving HUD‐VASH services were given a description of HF's core components in lay terms, queried about their perspectives on the practice, asked to describe the housing process, services received, relationship with staff, and unmet needs; these interviews (~30 min/each) were audio‐recorded and transcribed verbatim. Recruitment of veterans ended once research staff felt we had reached saturation of qualitative themes.

TABLE 2.

Participant demographics

Homeless‐experienced veterans interviewed
Category	Percent (n = 31)
Male	100 (31/31)
Psychiatric diagnosis	58 (18/31)
Depression and/or anxiety	39 (12/31)
Bipolar disorder	10 (3/31)
Psychotic disorder	10 (3/31)
Substance use disorder	52 (16/31)
Alcohol use disorder	29 (9/31)

Providers interviewed across six HUD‐VASH teams
Category	Percent (n = 51)
Social workers	61 (31/51)
Consumer providers (peer support specialists)	20 (10/51)
Nurses or nurse practitioners	16 (8/51)
Administrators	4 (2/51)

Six HUD‐VASH teams surveyed
Role	Percent (n = 40)
Social workers	50 (20/40)
Consumer providers (peer support specialists)	25 (10/40)
Nurses or nurse practitioners	25 (10/40)

A similar protocol was followed for qualitative interviews with eligible providers, who included HUD‐VASH teams and administrators. Efforts were made to recruit providers with diverse backgrounds and roles. S.H. and A.F. recruited providers (n = 51) out of a pool of 122 at HUD‐VASH monthly meetings and via email and phone. S.H. and A.F. conducted the majority of individual interviews, while coauthors R.O. and J.G., experienced researchers, conducted five interviews with providers. No participant was lost to the study after consenting. Table 2 describes provider characteristics. Interviews were conducted in person at VA offices or over the phone. Across all seven teams, providers were asked to describe HF sustainment and contextual factors that supported or impeded sustainment, based on an interview guide structured by the integrated sustainability framework. Interviews with providers included questions about changes to HF practice over time, structural challenges, and community partnerships, as well as strengths and needs; interviews (~45–60 min) were audio‐recorded.

Five coauthors—S.G., A.F., J.G., S.H., and R.O.—used a team‐based template analysis to code data using a priori code list derived from the integrated sustainability framework. ³¹ Major conceptual categories included (1) outer contextual factors, (2) organizational processes, and (3) practice sustainment; across these categories, there were 30 additional subthemes and subcategories. The team created operational definitions of subcategories, for example, “sociopolitical context,” then each team member coded a subset of interviews based on sustainment factors and notes on potential codes and trends in interviews. Qualitative data analysis software (ATLAS.ti 9) was used to organize and code data. Two authors (E.F. and A.F.) discussed notes on emergent patterns and processes, finalized a codebook, then assessed the transcripts to ensure all codes and themes were applied uniformly and to eliminate repeating ideas. E.F. selected participant quotations to illustrate themes and searched the entire data set for disconfirming cases.

2.4. Sustainment survey

Six of seven HUD‐VASH teams were recruited to provide self‐reported data on HF sustainment via a 1‐h survey. Groups of three to six providers from each participating team were asked to discuss and respond to a 46‐item Housing First survey, a validated measure of HF fidelity. ³² The survey was performed at VA offices. Audio of the group responses was recorded; spreadsheets summarizing their reported results were created.

As we captured this fidelity measure 5 years after intensive HF implementation efforts, we considered it a proxy for sustainment. This ordinal‐scale survey measures sustainment to five core components of HF: (1) housing process and structure, (2) separation of housing and services, (3) service philosophy, (4) service array, and (5) team structure. ³² The quantitative sustainment measure was calculated as described by its developers. ³² Scores ≤16 were considered low sustainment, while scores ≥28 indicated high sustainment (Todd P. Gilmer, PhD, email communication, November 2017). Fidelity scores from participating HUD‐VASH teams were compared.

2.5. Data integration

Quantitative and qualitative data were integrated during the analysis phase of this study. ³³ Researchers E.F. and J.G. independently triangulated results from the sustainment measure with qualitative findings via a convergent coding matrix and compared the data sets for points of divergence, partial agreement, convergence, and silence. ³⁴ , ³⁵ In subsequent meetings, they compared and synthesized results, then interpreted points of partial agreement and divergence, in accordance with triangulation protocol outlined by Farmer. ³⁵

3. RESULTS

3.1. Housing First sustainment

Overall, HUD‐VASH teams reported sustained HF, with high sustainment scores across all but one team and little variability across sites. Sustainment to the HF fidelity domains of housing process and structure, separation of housing and services, and service philosophy (see Table 1) received a mean score of 35 ± 1.26, indicating high sustainment. The domains of service array and team structure received a mean score of 29.58 ± 3.87, also consistent with sustainment.

Highest sustainment was measured in service philosophy and separation of housing and services (e.g., lack of treatment mandates). Lowest sustainment was found in the domains of housing process and structure (e.g., time to housing) and team structure (e.g., client‐to‐staff ratios and frequency of client–provider contact): Consistent with lower HF fidelity was an average time from enrollment to permanent supportive housing of 6 months, and an average veteran‐to‐staff ratio of more than 36 veterans to one full‐time employee (excluding physicians, nurse practitioners, and administrative support). In the domain of housing structure, two teams reported a low percentage (0%–14%) of veterans paying 30% or less of income toward rent, while all other teams reported 85%–100% of veterans. Two teams reported discharge from HF after a certain number of attempts to engage rather than maintaining contact with all enrolled veterans, which represents lower HF fidelity in the domain of separation of housing and services. Three teams reported goals that are chosen by HUD‐VASH teams rather than a veteran‐centered approach to goal setting; this corresponded with lower fidelity scores on the domain of service philosophy. As a whole, sustainment results indicated a lack of program drift and fidelity to HF in continued practice implementation.

3.2. Outer and inner contextual factors

Figure 1 depicts the integrated sustainability framework in relation to HF sustainment in the HUD‐VASH program at VAGLA. Table 3 summarizes the impacts of outer and inner contextual factors on HF core components, by domain, and HF sustainment. We detail these impacts below.

FIGURE 1.

Integrated sustainability framework adapted for Housing First sustainment at VAGLA's HUD‐VASH program

TABLE 3.

Impacts of outer and inner contextual factors on HF core component domains and sustainment

	Impact on HF core component domain(s)	Impact on HF sustainment
Outer contextual factors
Federal mandates to end veteran homelessness	Housing process and structure	Staff found it difficult to minimize time to housing, due to high demand and low housing availability.
Competitive rental market in Los Angeles	Housing process and structure	Because affordable housing was often located in under‐resourced areas, veterans felt that they lacked housing choice and felt unsafe at their housing location. This in turn affected the perceived benefit of HF.
Traffic and urban sprawl in Los Angeles	Team structure	Contributed to a limited number of in‐person visits with residents.
Inner contextual factors
Leadership turnover	Housing process and structure	Leadership changes resulted in miscommunication issues among HUD‐VASH teams and contributed to challenges associated with navigating VA bureaucracy and a lack of advocacy for additional resources, which in turn slowed residents' time to housing.
Organizational resource limitations	Housing process and structure; team structure	A lack of equipment, vehicles, and space limited HUD‐VASH teams' ability to operate, slowing residents' time to housing and limiting in‐person visits with residents.
Staff turnover	Housing process and structure; team structure	Limited staffing and high turnover resulted in higher staff‐to‐provider ratios, which slowed time to housing and number of in‐person visits with residents.
Processes
Community partnerships	Service array	Improved community partnerships enabled HUD‐VASH teams to refer veterans to needed services not provided by the VA.
Characteristics of the interventionists
Training and supervision on HF practice	Service philosophy; service array	New employees lacked sufficient training and supervision on service philosophy, strength‐based approach, and local community services.
Characteristics of the intervention
HF adaptations to fit the context and structure of HUD‐VASH at VAGLA	Team structure	Adaptations to team structure facilitated additional in‐person visits with veterans.
Sustainment
Perceived benefits of HF	Service philosophy	A high perceived benefit reinforced the need for low‐barrier rapid, permanent supportive housing.

3.3. Outer contextual factors

External factors impacting HF sustainment included pressure to achieve practice success, modulated by an increasingly competitive housing market. ²⁹ , ³⁰ Providers identified a disconnect between federal funding and mandates to house veterans and the structural challenges they saw on the ground that thwarted this goal. More specifically, they noted intense federal and local pressure to improve housing outcomes, which failed to fully account for housing scarcity and high rent as major barriers to practice success. One provider remarked, “The [federal] expectations are unrealistic … They want to try to house all these Veterans. You've got a 2% availability rating for apartments versus thousands of Veterans who still need to be vouchered—1300 still.”

Many providers detailed limited options for affordable housing in Los Angeles. These providers asserted that such options could only be found in under‐resourced areas that inhibit veterans' recovery and prevent them from making meaningful choices in their housing location. One provider noted, “It breaks my heart, but we're housing them in not‐so‐safe areas … These Vets, they're trying to get over their addictions, they're trying to get over their trauma, and then we put them in the ghetto.” Similarly, one veteran participant stated, “[HUD‐VASH] put us in drug addict areas and with drug people and thieves.”

3.4. Inner contextual factors

3.4.1. Leadership turnover

During the first few years of HF practice, significant changes in leadership led to miscommunication, as well as interpersonal conflicts within HUD‐VASH teams. Some also related leadership issues to a slowness to adopt HF among providers, also citing challenges associated with implementing culture change at the VA, an organization with significant bureaucracy. Some providers also described a lack of sufficient training and education to implement HF as barriers to practice sustainment. However, as the practice matured and new leadership was hired, providers reported supportive supervisors and administrators as an integral asset to their ability to operate a program aligned with HF's key components.

3.4.2. Organizational resource limitations

Limited resources contributed to a challenging work environment, as did transportation issues associated with the metropolis's sprawl and high traffic. Providers frequently noted long distances to VA facilities, extensive time spent transporting veterans to receive VA services, and a limited number of VA vehicles as major barriers to practice operation. Limited space and resources at VA offices included a lack of parking, a lack of office equipment, and insufficient physical space. Providers also noted difficulties associated with service linkages to other VA services, for example, strict eligibility criteria, limited psychosocial rehabilitation for substance use disorders, poor care coordination, long waiting lists, a lack of data sharing, and the absence of transitional housing options. In turn, these issues lengthened the housing process.

3.4.3. Staff turnover

Providers indicated that long wait times for staff hire and clearance resulted in significant staffing shortages; with inadequate staff, teams were not able to visit veterans at their residences as frequently as needed, which worsened housing stability. In turn, higher caseloads from staff shortages led to burnout, which contributed to turnover. On the issue of burnout, one provider noted:

There's this ongoing cycle of people losing housing because they can't get enough support after they've been housed, versus all the energy going towards meeting the numbers to get the Veterans housed and off the street … I truly felt like I was running around sometimes like a chicken with my head cut off, just one task to the next to the next. [It is] too much.

Most veterans interviewed appreciated regular home visits by their team members and highlighted the importance of individualized support and specialized case management for housing maintenance. However, they also noted high caseloads and turnover as major practice challenges. One veteran said, “I just believe they [HUD‐VASH] need to work more on … hiring more people, so the workload is manageable. So the workers can be efficient.” When asked about ways to improve the practice, another veteran stated:

I wish you guys [HUD‐VASH] wouldn't work these poor guys [on HUD‐VASH teams] to death—too many caseloads … I know the VA is well‐known for that: overwork the doctors, overwork the counselors … Man, if only they could thin out the caseloads, so they could get to people that need the attention.

3.5. Processes

3.5.1. Partnerships

Providers noted improved community partnerships to be integral to practice operation. Following HF implementation, community engagement became a priority, and VA representatives participated in citywide efforts to streamline and coordinate homeless service provision. Providers noted increased ties to various VA services and community organizations as key to HF sustainment. One provider described the importance of community partnerships, saying, “We can't operate like an island. We are very interdependent upon … different agencies to get our job done.” Veterans also recognized the benefits of their team linking them to additional services, and many indicated they had utilized community organizations to locate housing; pay security deposits; find employment; and acquire furniture, food, emergency funds, and other necessities. For some veterans who had experienced chronic homelessness, these resources served as a lifeline while they habituated to housed life.

3.5.2. Training/supervision

Given the expansive areas served by HUD‐VASH teams, one provider in an outlying service area stated the need for greater inclusion in all‐staff meetings and support for staff: “It would be, really, a morale boost, and it would also be a retention benefit if we were included more in … team meetings and the trainings, just to feel more connected to what's going on.” Additional findings suggest the need for additional training, mentorship, and supervision of staff, to ensure adherence to HF service philosophy and facilitate knowledge‐sharing about internal processes and community‐based resources to support veterans.

3.5.3. Adaptation: Changes in multidisciplinary teams

The team‐based service provision approach changed significantly over the first 5 years of HF implementation. Initial specialty teams included an assertive community treatment team to serve “high needs” veterans, as well as an intensive case management team to serve veterans with identified “moderate needs.” Other teams served lower‐acuity veterans, and one team was dedicated to the intake and assessment of veterans. However, by the time the study was conducted, the specialty teams had adapted their staffing structure and targeted population to the extent that they were effectively dissolved as specialty teams and re‐tooled as teams that worked with veterans of all acuity levels, with each team assigned to a service planning area. Assertive community treatment teams and intensive case management teams are common within the HF model, ³⁶ , ³⁷ but given turnover, heavy traffic, and urban sprawl, administrators decided it was more practical to build teams' capacity to conduct outreach to all veterans within a set region. Providers agreed that the current approach to mixed‐acuity team‐based work improved morale and enabled more effective communication within and across teams.

3.6. Characteristics of interventionists

Given high staff turnover, management struggled to fully orient and train new employees in HF. Likewise, new hires often included recent graduates of master's level social work programs, and these employees were often unfamiliar with VA‐specific resources and processes, community resources, and the homeless‐experienced veteran population. When asked about ways to improve HF practice, providers brought up the need for additional training, supervision, and mentorship. One provider remarked, “They didn't prepare us for this [conflicts with Veterans] in school, and the VA is not preparing us either … It's just like trial and error for us … So they [the VA] really should invest more money into training us.” HUD‐VASH teams also noted the need for greater role clarity, detailed guidelines, and protocol enforcement to support practice maintenance.

3.7. HF characteristics

The perceived need for HF practice was high among all stakeholders. When asked about the practice's main strengths, providers discussed veteran‐centered support, low barrier entry, and rapid housing. To the latter point, an administrator remarked:

[The HUD‐VASH housing subsidy] offers a resource that is otherwise unavailable in LA … Waiting lists for regular [vouchers] are 10 years or more. So this is a great resource for Veterans to obtain [housing] very quickly. Somebody can enter our program and have a [housing subsidy] within 2 weeks of being admitted. It's unheard of [outside VA].

Veterans indicated the need for rapid permanent supportive housing, and found that HF enabled them to feel a sense of stability and safety and improve their health. Some indicated that housing enabled them to work on other goals, such as finding employment or reconnecting with family. One veteran stated, “I think it's one of the best setups I've ever seen, because it's so quick. I mean, anybody that goes in and does the paperwork, you're going to have your [housing subsidy] within two weeks. You can't do that anywhere [as a civilian].” Another veteran remarked, “I'm not out in the street where you're vulnerable to everybody … Having the security [of housing], it gives you self‐esteem, so it made me want to do more positive things with my life.” Veterans also noted wraparound services (e.g., transportation to medical appointments, service linkage, and peer support) to be major assets of HF, along with specific aspects of their housing unit and building.

3.8. Continued benefits

Veterans reported that HF gave them a sense of reassurance and security. One veteran explained, “I know that if something went wrong with [my housing], I could be re‐enrolled in [the program], because I'm a Vet. I have disabilities. I know that, and I'm not going to be perfect again, but I feel I'm strongly stabilized and capable.” Other veterans felt as though they could set goals and plan for the future. To this point, one veteran stated, “I just really appreciate everything. HUD‐VASH got me out of the streets, and now that I have a foundation … I'm looking, maybe, to save up and buy a house.”

4. DISCUSSION

High fidelity scores across nearly all core domains of HF practice, across HUD‐VASH teams at VAGLA, suggest sustainment of this practice 5 years after intense implementation support ceased in the nation's largest HUD‐VASH program. ²⁰ This study identified complex relationships among multilevel factors that affect HF sustainment and highlighted the need for strong leadership to promote practice sustainment. Adaptations to HF in response to outer contextual factors and organizational capacity can help programs maintain fidelity while allowing for necessary flexibility in service provision.

Although high fidelity scores were achieved 5 years after implementation, a number of factors limited the effectiveness of the programs. Outer contextual barriers to HF sustainment (e.g., a highly commoditized housing market, metropolitan traffic, and sprawl) impacted inner contextual factors (e.g., limited organizational resources), which slowed housing processes. For example, HUD‐VASH teams were expected to help veterans with housing searches, but often lacked a sufficient number of VA vehicles; when they did have transportation, the great distance from their offices to potential housing meant that much of their workday was spent driving from place to place. These factors slowed time from enrollment to housing.

An adaptation to assign HUD‐VASH teams based on service planning areas, rather than veteran acuity, streamlined operation, positively impacting the fidelity domains of housing process and structure, as well as team structure (with respect to contact frequency). It shares similarities to the Dutch practice of using a single team to provide assertive community treatment and intensive case management levels of services, an approach often termed “Flexible Assertive Community Treatment.” ³⁸ In contrast, a national HUD‐VASH study suggested the potential value of acuity‐based multidisciplinary teams, to provide more targeted support for high‐need veterans. ¹⁵ Given the lack of consensus in the literature, this adaptation may be region‐specific and non‐generalizable, yet it demonstrates how a regional adaptation can enhance fidelity and sustainment.

This study found that a major inner contextual factor that impacted HF sustainment was leadership change, which made it difficult for leaders to advocate to upper management on behalf of HUD‐VASH teams. Meanwhile, HUD‐VASH teams managed sizable caseloads with limited organizational resources. This working environment contributed to staff burnout and turnover, while limited training and supervision resulted in a lack of role clarity among recent hires. This impeded care coordination. Likewise, other studies on HF implementation and sustainment at VA indicate the value of middle management in communicating practice needs to upper administration, in order to address staffing issues, acquire organizational resources, and foster collaborative teams. ¹ , ¹⁵

Perceived benefit of HF was high across all stakeholders, yet modulated by outer and inner contextual factors. While HUD‐VASH enabled veterans to quickly acquire housing subsidies, stakeholders reported outer contextual barriers to using them, including a competitive rental market. This lack of housing choice in HUD‐VASH intersects with health and safety concerns about housing units that accepted HUD‐VASH participants, particularly in under‐resourced areas. In some cases, stakeholders believed housing location facilitated substance abuse and limited veterans' ability to maintain housing. This finding is consistent with results from studies on HF fidelity, which similarly note housing choice as difficult to achieve in gentrified housing markets where potential landlords are hesitant to rent to veterans. ¹⁴ , ³⁹ , ⁴⁰ Meanwhile, inner contextual facilitators of HF included HUD‐VASH‐supported collaborations to identify and secure community resources. Community partnerships contributed to perceived benefits of HF among veterans, who often listed linkages to furniture donations, security deposits, and other services as important to their housing maintenance. These efforts expanded the service array of community‐based support. This finding is consistent with initial challenges associated with the rapid expansion of HF at HUD‐VASH sites, indicating the need for coordination with community agencies. ¹

As a whole, these findings confirm practice sustainment while highlighting opportunities for capacity building as the program matures. While outer contextual factors such as competitive rental markets will likely remain persistent challenges to HF sustainment, adaptive responses to the design and operation of HUD‐VASH teams, that is, HF practice adaptations, as well as increased coordination with housing agencies and community organizations, could support improvements, such as reductions in time to housing. ¹ , ¹⁴ , ³⁹ Moreover, they suggest the necessity of organizational leadership, additional resources, and strong community partnerships in HF sustainment. ¹⁴ , ¹⁵ , ¹⁶

The integrated sustainability framework provides a comprehensive set of multilevel factors that interact and respond to changes in outer and internal contextual factors to HF sustainment. ²⁸ Analyzing data using this conceptual framework revealed the most salient issues, as described by VA stakeholders. This approach enabled our team to identify relationships between factors and describe key facilitators of and barriers to practice sustainment. In turn, continued research on HF sustainment could enable a deeper understanding of change over time and ways in which regional adaptations can improve evidence‐based practices. ¹⁴

This study had limitations. First, we used an observational design, and obtained data from a relatively small sample of providers and veterans, within a single VA health care system. As baseline fidelity was not captured during initial intensive efforts to implement HF, we do not have objective points of comparison to measure fidelity and sustainment over 5 years. Fidelity was assessed via surveys of HUD‐VASH teams, rather than direct observation in the field and other data sources to track program metrics, such as veteran housing and health outcomes. Potential modifiers to this method included unknown power dynamics between team members and a positive or negative bias in reporting outcomes. Lastly, the integration of our qualitative and quantitative findings is limited by our use of provider interviews to calculate the fidelity scale. Despite these limitations, this study provides a useful starting point for evaluating key regional and organizational issues that impact HF sustainment several years after implementation, to ensure practice adaptations do not lead to program drift but enhance HF sustainment.

These findings contribute to the emerging literature on the sustainment and adaptability of HF within specific regions and systems of care. ¹⁴ Given HUD‐VASH's adoption of HF as an effective practice, it is crucial that practice implementation remains responsive and adaptive to situational factors that impact sustainment and, in turn, affect the health and safety of veterans. Characterizing HF sustainment can identify components of HF practice that have the potential to enhance future HF implementation, fidelity, and sustainment efforts.

CONFLICT OF INTEREST

No potential conflicts exist.

Fletcher EH, Gabrielian S, Flynn AWP, et al. Stakeholder perspectives on sustainment of Housing First in a VA permanent supportive housing program. Health Serv Res. 2022;57(2):374-384. doi: 10.1111/1475-6773.13904