Implementing a hospital-based case management intervention for people experiencing homelessness: the navigator program

Abstract

Background

People experiencing homelessness (PEH) have worse health than the general population, and higher rates of hospitalization. The transition period after discharge from hospital is often challenging for PEH, in part due to loss to follow-up, competing priorities, housing instability, and the absence of a primary care provider. In-patient hospital stays represent a window of opportunity to intervene and connect with patients, supporting them to stay in hospital and complete their treatment plan, identify and address their social needs, and support their transition of care into the community. This qualitative study explores supports and challenges to the implementation of the Navigator Program, a hospital-based critical time intervention that supports PEH during their hospital stay and after discharge into the community.

Methods

We interviewed 35 participants (homeless outreach counsellors working on the program, hospital physicians and staff in the implementation setting, community service providers, and the implementation team) and conducted 130 h of non-participant observation. Analysis used the Framework Method and the Consolidated Framework for Implementation Research to highlight the barriers and facilitators to implementation.

Results

A core aspect of successful implementation and program uptake was that all participants saw a need for the program. The flexible approach to model design and implementation was an essential approach to program development that adjusted to the implementation setting, while leaving room to create more systems and structures as the program progresses. Implementation also relied on clear approaches to attaining buy-in from all stakeholders, done through a mix of formal and informal approaches. Operating as a hospital-based program was essential for successful implementation, supporting team-building among care providers in both the healthcare and social service sectors, which can lead to improved patient care coordination.

Conclusion

The implementation of programs addressing complex social and health issues can contribute to its success or failure. In this study, we discuss the effective implementation approaches of the Navigator Program, as well as lessons learned. This study provides practical and helpful strategies for implementing similar programs in hospitals across Canada, and in countries with similar healthcare system structures.

Background

People experiencing homelessness (PEH) have disproportionately worse health and are at greater risk for mortality than the general population [1,2,3,4,5], yet face numerous barriers accessing primary and preventative healthcare [1, 6,7,8]. This can result in treatment delays, worsening health status, high rates of hospital admission and emergency department visits, and longer in-patient hospital stays [1, 9,10,11,12]. In 2022–2023, 30,000 hospitalizations in Canada were attributed to people experiencing homelessness [13, 14]. These hospitalizations were twice as long (15.4 days vs. 8.0 days) and twice as costly (CAD $16,800 vs. $7,800) when compared to the national average [13]. PEH may experience suboptimal in-hospital quality of care, marked by unfeasible discharge plans and inadequate care coordination with community service providers (CSPs), making recovery a significant challenge [12, 15,16,17]. After discharge to the community, recovery challenges are further compounded by inability to carry out the care plan, competing priorities, housing instability, and the absence of a primary care provider [18,19,20,21]. All combined, these challenges can lead to a higher rate of hospital readmissions following initial hospitalization. One urban downtown Toronto hospital observed a 27% readmission rate within 90 days of discharge [16], with informal supports, communicating discharge plans to primary care physicians (PCP) and having an active case manager associated with lower risk of readmission. Other studies have reported 90-day readmission rates ranging from 21 to 40% [22, 23]. By comparison, housed individuals had lower 90-day readmission rates (21 to 23%) after adjusting for numerous clinically relevant factors [23].

In-patient hospital stays represent a window of opportunity to intervene and connect with patients, supporting them to stay in hospital and complete their treatment plan, identify and address their social needs, and support their transition of care into the community. To meet these goals, various case management programs tailored to the needs of PEH have been implemented in hospitals, especially for frequent users of healthcare services [24, 25]. Critical Time Intervention (CTI) is one predominant short-term case management model widely adopted in healthcare. Research indicates that CTI programs can help effectively reduce hospital readmissions and enhance healthcare navigation for marginalized populations [26, 27]. Specifically, CTI has been shown to increase outpatient service use and promote continuity of care for PEH [28,29,30,31,32]. Despite CTI’s proven effectiveness, local adaptations in different contexts, and implementation approaches and outcomes are less understood and require further research [33].

Evaluating the implementation of an intervention is important to provide a feedback loop that helps researchers assess an intervention’s effectiveness within a specific context [34]. Successful implementation relies on intervention deliverers taking up the intervention as intended, and therefore requires a strategy that considers peoples’ motivations, actions, intentions, and feelings about the program [35]. Yet, research on the implementation of interventions that provide case management support for PEH during and after hospitalization is limited [31, 32]. Evaluations of programs providing case management post-discharge [36, 37] and housing support while hospitalized [38] to homeless or unstably housed patients report common barriers and facilitators of implementation effectiveness, presented in Table 1. The present study adds to this body of literature by examining the implementation of the Navigator Program, an adapted CTI that provides case management services to PEH during their hospitalization and for approximately 90 days post-discharge [39]. We aim to better understand how aspects of the implementation process, the implementation setting and the broader context shaped how the program was implemented (e.g. the training and resources necessary to achieve full implementation), aspects of the implementation process that were beneficial or challenged the program’s operations, as well as improvements to be made for future implementation in subsequent settings [40].

[IMAGE OMITTED: SEE PDF]

Methods

Study design

This present study is part of a broader pragmatic, qualitative process evaluation of the Navigator Program, informed by the UK Medical Research Council (MRC) guidance on process evaluations of complex interventions [40]. The process evaluation focused on implementation, mechanisms of change, and context. This study reports on findings on factors that shaped implementation of the Navigator Program. By implementation, we refer to a series of intersecting processes that aim to get an intervention or program into use within a specific setting [35]. The process evaluation was conducted concurrent to a pragmatic Randomized Controlled Trial (RCT) of the Navigator Program, described in detail elsewhere [39].

Study setting

This study took place in Toronto, Canada, where approximately 12,000 people experience homelessness nightly (although this number is likely an undercount). An estimated 22, 000 individuals were served by Toronto’s Shelter Support Services in 2023 [42], and the shelter system, while continuously increasing the number of beds available, is regularly at 95 − 100% capacity [43].

The Navigator Program was designed and implemented at an urban academic teaching and community hospital in Toronto. In the fiscal year 2023–2024, administrative data showed that this hospital had 903 hospital admissions related to 638 unique individuals experiencing homelessness, and had 10,190 visits by 3,527 unique individuals in the Emergency Department– a volume of unhoused patients greater than all other Toronto hospitals. The Program was designed in collaboration with hospital physicians and staff, community partners and people with lived experience of homelessness [39]. Since inception in 2019, the Program has provided services for over 900 patients. Starting with an administrative and management support team and one Homeless Outreach Counselor (HOC), it currently employs three full-time HOCs and one part-time HOC who are responsible for creating strong links between patients and community services.

HOCs meet patients in the hospital upon admission, support them to stay in-hospital for their treatment, and then work with their care team to make their discharge plan a reality. They follow patients into the community, supporting their health- and social-related needs. HOC activities include connecting patients to a primary care provider, providing transportation support to follow-up medical appointments, coordinating wound care, obtaining assistive devices, helping clients file their taxes, acquiring identification and health insurance or replacing lost cards, filling out housing applications, and connecting them to longer-term case management, among many other activities. The Program provides services for approximately 90 days, with flexibility to extend support for longer if needed. A discretionary fund allows HOCs to use car hire services (e.g. Uber) to support transporting clients to medical appointments, purchase medical equipment and clothing, provide clients with cellphones, pay for medication until their insurance is activated, and cover other health- and social-related expenses.

Participant sampling and recruitment

Participant groups included the implementation team (researchers and hospital management), HOCs, hospital staff and physicians, and community service providers who interacted with the Navigator Program. Purposive sampling was used to identify potential participants. At the time this study was conducted, there were two HOCs in the intervention and four people involved in the program implementation who all agreed to participate. We worked with the HOCs to identify hospital physicians and staff, and community service providers whom they suggested would be best situated to comment on their experiences working with the Program. We recruited hospital physicians and staff from different inpatient services and units in the hospital where this program was implemented (except for psychiatry and obstetrics as the program does not currently support these services), and community service providers from a number of different sectors including emergency shelters, housing, addictions, mental health, and primary care. During analysis, we interviewed two more HOCs who had been hired onto the team since the initial participant recruitment, to gain a sense of developments in the program. Individuals were eligible if they were 18 years of age or older, capable of providing consent, and spoke English. Individuals who expressed interest in participating were taken through the consent process and all provided written or verbal (if the interview was done on the phone or video) consent to participate.

Data generation

Two methods of data generation were used: semi-structured interviews and non-participant observation (NPO). Semi-structured interviews ensure that all necessary concepts are covered but also allow for the flexibility of unanticipated topics, concepts and follow-up questions to be explored. From December 2022 to September 2023, interviews were conducted with participants in person or over zoom, at their preference. Interviews explored reflections of the hospital setting and the implementation decisions and process, experiences working with the Navigator Program both within and outside of the hospital, and opinions about how this program functioned (see supplementary material interview guides). Follow-up interviews with the two HOCs who joined the team after implementation discussed their experiences with onboarding, if they felt supported, ongoing challenges, and other experiences to help understand the impact of adjustments made during implementation. Interviews were audio-recorded, transcribed verbatim by a professional transcription service, audio-checked for accuracy and uploaded to Dedoose.

NPO is a process of observation of participants and the intervention setting without actively participating, and can be “useful for capturing finer details of implementation, examining interactions between participants and intervention staff, and capturing aspects of the ‘spirit’ of implementation, rather than just the mechanics of its delivery” [40]. The study lead/first author and research coordinator conducted a combined 130 h of NPO between November 2022– February 2023, shadowing the HOCs in the hospital and community, and attending program meetings. Field notes recorded observations on how the program was operating and helped contextualize interview data.

Data analysis

We employed the Framework Method as a tool for data management, guided by the Consolidated Framework for Implementation Research (CFIR), to explain barriers and facilitators to implementation effectiveness [44]. The Framework Method follows five steps: transcription, familiarization with the interview, coding, developing an analytic framework, and applying the framework [45]. The CFIR, which was used to analyze what shaped or impacted program implementation, consists of five domains: Innovation (the thing being implemented), Inner Setting (the setting in which the innovation is implemented, such as the hospital), Outer Setting (the setting in which the Inner Setting exists, such as the healthcare system), Individuals (the roles and characteristics of individuals), and Implementation Process (the activities and strategies used to implement the innovation) [35, 44]. Each domain has a number of constructs to guide evaluation; we selected constructs most relevant for this study. For a worked example of the Framework Method see Table 2.

[IMAGE OMITTED: SEE PDF]

Following the Framework Method, transcripts were read by three research team members (JJ, OD and DM) to familiarize ourselves with the data. Analysis of participant interviews and NPO field notes happened simultaneously, moving back and forth between the two data sets. For interviews, we independently conducted initial coding on three transcripts, informed by the CIFR domains and constructs as well as codes emerging from the data. This first stage of coding took place in Dedoose Version 9.0.107. CFIR codes were applied deductively to the data; pieces of text that were coded inductively were grouped under the relevant CFIR construct or domain parent code. After ensuring our coding approach was in alignment, we agreed on the analytic framework and coded the remainder of the transcripts. Next, two team members (JIRJ and DM) applied the framework matrix (see Table 2 for an example) [45]. For each participant, data assigned to each code were reduced to summary statements with illustrative quotations. This enabled analysis within and across participants and codes, allowing us to map connections and relationships within and across CFIR domains. We therefore moved from description of the data to developing analytic themes that, conceptually, extend across and beyond the CIFR domains. During this stage of analysis, higher level codes and, eventually, analytic themes were applied to the NPO field note data. This allowed us to compare and contrast interpretations of the interview data with NPO field notes, contextualizing the interview data and either making adjustments to interpretations or expanding on ideas as needed. Throughout the analytic process we recorded analytic ideas, interpretations, and conceptual developments in memos and used these memos as a starting point to help generate themes.

Results

We conducted a total of 35 semi-structured interviews with four members of the implementation team, four homeless outreach counselors (HOC), 14 hospital physicians and staff, and thirteen community service providers (CSP). See Table 3 for participant demographics. Analysis and interpretation led to the development of five main themes: (1) recognizing a need for the program; (2) balancing flexibility and structure; (3) strong buy-in and a clear benefit to all stakeholders; (4) hospital embeddedness and feelings of being rooted; (5) and, team building processes within and outside the hospital.

[IMAGE OMITTED: SEE PDF]

Recognizing a need for the program

During interviews, participants indicated that before the Navigator Program, their ability to support PEH in hospital and during their discharge was difficult. Across groups, participants highlighted a gap in services when discharging PEH from hospital to the community, and a need for more supports both within and outside the hospital (inner and outer settings). This was partially attributed to the high number of unhoused patients the hospital cares for, combined with the lack of time hospital care teams had to build relationships with patients. Participants discussed experiencing pressure to discharge patients as soon as they were medically optimized, leaving little time to address the overall health and social needs of PEH. Hospital physicians and staff discussed their constraints when considering patients’ health and social needs post-discharge:

There’s only so much the hospital can do. I think we focus more on medical needs, but there’s so much more, like, the finances, the medication, the benefits, housing, you know, even food, transportation, getting a phone. All these people, you know, don’t do well or come back even sicker because they don’t have the basic needs established. (Hospital_3)

This participant captures an understanding that was held by many, that when a patient’s needs are not met in the community they may return to hospital. Other participants explained that the health system and out-patient services were confusing and difficult for patients to navigate (e.g. keeping track of follow-up appointments with community healthcare providers) making it hard to complete their discharge plan. A member of the implementation team explained that they were often uncertain about whether patient discharge plans were feasible:

I just knew that in most of those cases for patients that are homeless, it was like we were rolling the dice. And there was a very high likelihood that many if not most of the things we said to be done after discharge would just not happen. (Implementation_1)

Other members of the implementation team reflected that clinical teams had been struggling for a while: “our [clinical] team identified the gaps before and needed help. And we couldn’t fulfill all of the support that sometimes our patients needed,” (Implementation_3). Participants felt that hospital workers have, “a lot of interest and frustration… people knowing that they’re not serving this population very well. And being really interested in anything that will help them do that better,” (Implementation_4).

When asked if there was a need for a program like Navigator, all participants said yes.

Sometimes, you don’t even know when you need something until you have it, and then when it’s not there, you’re like how did we ever live without this? I don’t know how we ever lived without HOC_2 before they enlisted, like they just have the answers and solutions to all the problems. (Hospital_14)

The community perspective was that, “hospitals don’t know where to put people,” and so they would often discharge to destinations where, “it’s not necessarily [an] appropriate admission,” (CSP_3), such as an emergency shelter. However, there are few appropriate options in the community. Many spaces are not physically accessible, for instance: “We have a lot of stairs as well, we don’t have escalator or elevator” (CSP_2). This limited options for discharge destinations, and increased the likelihood that the only option to hospital physicians and staff was to discharge patients to spaces not able to meet their needs during recovery after hospitalization.

Community service providers also saw a need for unhoused patients to be connected to housing and other administrative supports (e.g. getting ID, completing taxes to qualify for benefits), and they were struggling to meet the demand. Participants noted that the current condition of the shelter system was challenged to support patients from hospital. They have staff “who are burnt out, who are overworked, they don’t have the proper support” (CSP_ 11). Another participant (CSP_4) recounted a story about a patient with cancer staying at a shelter hotel. After chemotherapy, the patient was tired but the elevator didn’t always work and he was on the 8th floor, and he needed to change his diet, which the shelter could not accommodate. As the participant clearly states, “the shelter is not built for patients.” CSPs were clear that the staffing and space resources in the community to support PEH after hospital discharge were insufficient.

Altogether, participants suggested there was a clear, recognized need for a program like Navigator.

Strong buy-in and a clear benefit to all stakeholders

Strong buy-in from individuals at different levels (frontline, middle level management, high-level directors, donors) of the inner setting (the hospital) was said to be crucial for facilitating successful implementation of the Navigator Program. Generating strong buy-in early in the program’s conceptualization was partly attributed to the program’s champion, who is a well-respected and trusted physician with extensive experience working with PEH both in the community and in the healthcare system. Participants reported that the reputation of the program champion, “preceded him with all of the work that he has done for homelessness in the city I think also speaks volumes too,” (Implementation_2). The trust stakeholders had in him was said to carry over to the program. When in the community, HOCs reported experiencing more acceptance of the program and their work because there was implicit trust in the program’s founder.

Strong support from hospital executives and funders/donors was also attributed, in part, to the simplicity, ease, and financial efficiency of the program. The program was said to be very marketable for fundraising purposes. Participants described it as a feel-good story: “we’ve got homeless patients and we’re helping them get the care they need with these wonderful people called navigators” (Implementation_1). Participants reported that there was also strong buy-in from mid-level management within the hospital. Achieving this support was said to be require more considered conversations, as it was asking these stakeholders to take on more work as supervisors for the HOCs. Study participants reflected that as the HOC role was unique and required flexibility, it was a significant ask for them to take on this responsibility. Therefore, getting buy-in from the managers on clinical service was said to be essential, and having them as a “strong supporter of the concept, and of the project” was “instrumental in making it work” (Implementation_1).

Buy-in from hospital physicians and staff was said to have developed over time, once the program was under way and they experienced the benefits to their own work. For instance, hospital physicians and staff expressed that the program alleviated moral distress they experienced when discharging people experiencing homelessness (see Cygler et al. [46] for more on this point):

I don’t think that I was doing many of the things that they do. So it wasn’t necessarily time that I was spending, but it was a mental angst that those things weren’t happening… It’s not that they save me 30 min, for example. It’s that they do things that allow me to focus on other parts that I know how to do. (Hospital_1)

For CSPs, a similar sentiment was expressed. As CSP_1 explained, working with HOCs “takes a bit of weight off of me.” There is a lot to coordinate after hospital discharge, “especially if someone has ongoing health needs,” and having someone else to help was seen as “a god send almost because I can trust that this will happen and that I have someone else to help me with this.” Of note, CSP participants reflected that HOCs had an important skillset that made them effective at their jobs, which increased their trust in the program. For example, CSPs reflected that HOCs seemed to clearly understand priority pathways for PEH to help connect them with a family doctor; had an understanding of the health services available in the community; and demonstrated familiarity, comfort, and “significant experience in supporting people experiencing homelessness… The Navigators know how to make people feel comfortable, know how to connect with people and understand what it means to put a low barrier in front of their services and meet people where they are at” (CSP_12).

In addition, CSPs discussed the flexible approach the program took to working with clients as a significant benefit to the program design:

The fact that they offer very low barriers and a flexible way of providing services and a non-judgmental approach to service. A trauma-informed approach to service can really come in from a deep understanding of the client’s perspective on entry into the service and a willingness to work with the realities of people’s lives. (CSP_13)

Cutting across all these levels of buy-in was the notion that the program fit well within the culture of the hospital. Part of the program’s acceptance was that it was seen to align with the overall mission of the hospital. As one participant explained, “the hospital prides itself on providing excellent care for patients experiencing disadvantage, so they see this as being very congruent with the hospital’s mission,” (Implementation_1). The program was also said to fit within the culture of the hospital, fit with the morals and values of the hospital employees: “I think we all chose to be at St. Michael’s because we like the nitty gritty. These are our people. These are the people that we serve,” (Implementation_2).

Balancing flexibility and structure

Participants reported that the implementation strategy was intentionally slow in developing the program model and rolling out the program. The program was implemented flexibly and iteratively to improve its chances of success. Starting with one HOC, the idea was to “hire someone, we’ll get it going, we’ll experiment and just see how it goes,” (Implementation_1). Participants suggested being flexible was important for adapting and responding to the context within and outside of the hospital. For instance, the program’s model includes connecting patients to community services, and therefore requires working relationships with CSPs. The implementation team decided that the HOCs were best placed to initiate these relationships, leveraging connections and trust they already had with CSPs. This understanding came from early work the implementation team did to understand the context, which included, “meeting with the discharge planners and Care Transition Facilitators and talking with them and really realizing that there was not a good connection between the hospital and the community,” (Implementation_4). As another participant explained, the community sees the hospital as “very rich… very well resourced and kind of maybe not in tune with what’s going on in the community,” (Implementation_1). Therefore, all participants felt it best for the HOCs to forge relationships between the program and the CSPs. There were numerous examples given of this happening, including this from CSP_2:

[I] met HOC_1 at a City event and we were actually sitting side by side… and we exchanged our business cards and contacts Then HOC_1 called two weeks later asking for shelter spaces, and we actually did have a space. So I helped with that process, we got the gentleman in and then that was the start of me working with the Navigator. (CSP_2)

The implementation team suggested that over time the relationships between HOCs and CSPs would grow, creating a more durable connection. Through our NPO and from interviews, we observed that this approach seemed effective, especially as the HOCs were said to have an excellent reputation in the community. We witnessed the familiarity between HOCs and staff at the different service agencies we would visit. Service staff knew the HOCs by name, knew the clients they were working with, and often provided any helpful information they had if we were looking for a specific client. However, this approach also meant there was less education (e.g. presentations) to CSPs about the program and what it could accomplish for clients. As a result, some CSPs said that they were not aware of the Navigator Program in general, but knew at least one of the HOCs by name. During interviews, some CSPs asked us to explain the program to them, even though they had supported shared clients and worked with the HOCs. Relying on informal relationships for the program to function well also added stress for some participants.

It’s mostly our informal relationships with people and agencies that carry this program. Which is silly, because this should all be formal, we should have formal pathways, because it’s so much more work when things are informal. It drives me crazy. (HOC_1)

Importantly, the flexible nature of the program included an easy referral process for hospital physicians and staff. For example, a decision made at the beginning of implementation was that the referral process could be informal. There was no paper work or forms to fill out. As participants explained, this made the referral process “seamless” (Hospital_10) and “so easy” (Hospital_11). Other participants discussed how fast the process was. As one of the social workers explained,

I call them [HOC] or email and they are pretty responsive, like they’re fast in responding and I usually always put like a little blurb about social history so that they know and the patient’s location, room number, etc., and then within that day I get a response. (Hospital_13)

The implementation team discussed an easy referral process as a key aspect of the program design. They suggested that the only way the intervention would work “is if essentially physicians have to do absolutely nothing to make it work well. Because if you say, oh this will really help you, but you have to do this thing. Then it’s like that’s a recipe for them to not do it,” (Implementation_1). During our hours of NPO, we observed the efficiency of the informal and flexible referral process. Hospital physicians and staff would email, text, call, or drop by the HOC office to notify them of an unhoused patient in the hospital.

Overall, while this flexibility supported the program development, it could also create some challenges. Although implementers felt that, “having somebody in the role and working with them to develop the role is actually something I think is really good that was part of the implementation plan,” (Implementation_4), challenges along the way made it clear there was a need for more structure. For instance, at the beginning of the program’s implementation, the HOCs were documenting their activities within a database; however they were not using standardized templates to create care plans with clients. Through NPO, we observed that HOCs identified a few needs at a time when working with clients, but did not necessarily go over the full spectrum of services they could provide. While done this way so as not to overwhelm clients, this approach made it difficult to assess the full spectrum of client needs early on, and then systematically identify client needs and their outcomes. During the program’s ongoing development, participants realized that a structured care plan template was necessary, to identify goals and outcomes of their work with clients. This documentation system was then co-developed with the input of the HOCs, which enabled it to adapt to their setting, context and workflow, as opposed to imposing a structure on them that did not necessarily reflect their work. Additionally, while having a flexible referral process was important at the beginning stages of the program, as referrals increased it became important to have a structured process for referring patients; one that remained easy and asked for little work on the part of the hospital physicians or staff, but allowed for a clearer organizing and documenting process.

Lastly, as reported in follow up interviews with more recent HOC hires, there was a need for a more formalized and supportive onboarding process for their role. They discussed that the current process included shadowing and asking key questions, but they would have benefitted from more structure and guidance. For instance, one HOC mentioned that having a sheet with all the important contacts, and who to call about what issue, would have been helpful. In addition, having more structured training sessions was said to potentially improve team building between the HOCs.

Hospital embeddedness and feelings of being rooted

A key design choice was to embed the program in the hospital, in contrast to most case management programs that are community-based and provide in-reach services at hospitals. All participants noted multiple benefits to embedding the program within the hospital. For instance, HOCs were hospital employees, had an office space within the hospital, and had access to patient medical records. This allowed HOCs to meet patients promptly upon hospital admission. As one participant (Implementation_2) explained,“[the] connection has to be made in the hospital,” since it could be a struggle to find patients after they had been discharged. Participants affirmed that HOCs needed to be “right on the floor” because referring out and having external people come in “just doesn’t work” (Implementation_4).

A key event that occurred at the same time as program implementation was the COVID-19 pandemic. Since HOCs were hospital employees, they maintained access to the hospital and to PEH in the hospital, while other community-based programs did not have access to their clients in the hospital. Conversely, however, hospital employees were restricted from going into spaces in the community, which created other challenges for finding and working with their clients once discharged from the hospital.

Being embedded within the hospital also meant that the HOCs were considered a part of a patient’s circle of care, which was also said to develop a sense of feeling “rooted into the hospital” (HOC_2). Participants explained that in order for community-based case managers to meet with patients or their care teams, patients had to consent to them being admitted as a visitor and then coordinate the visit. The implementation team discussed that the Navigator Program’s design circumvents this challenge. Since HOCs are hospital staff, they are “part of the circle of care, they don’t need someone to invite them in or they don’t need to come from somewhere else. They’re actually part of the care team… That makes the barrier to their involvement essentially zero. So, I think that was critical,” (Implementation_1).

Having the HOCs embedded within the hospital was also described as a benefit by community service providers. As CSP_1 explained, the HOCs provided them a hospital-based point of contact: “It’s just so nice to have someone on site there, who knows the context there.” CSPs felt that by being onsite the HOCs were better positioned to advocate for longer length of hospital stay or transfer to in-patient rehabilitation facilities. Although community service providers try to advocate for hospitalized clients, they reflected that “there is more power in being on site… and having those professional relationships,” (CSP_1). Another participant suggested that having HOCs embedded in the hospital improved communication and information sharing between the hospital and community: “HOCs will contact us and give us a really good overview of the person, what their care needs are, what their issues are,” (CSP_3).

Team building processes within and outside of the hospital

The idea of team building came up in three ways during interviews: team building between HOCs and the hospital physicians and staff, between HOCs and CSPs, and among the HOCs. Participants reflected that one of the main challenges they had anticipated during program implementation was ensuring that the HOCs were an integrated part of the hospital care team, and that their presence in the hospitals was accepted, while also maintaining their position as a complementary program. As one participant stated “our biggest thing was probably figuring out how they were going to fit in with our discharge planning team,” (Implementation_3). Decisions about the program design (e.g. having an office for HOCs on a hospital ward) were made to support integration into the hospital. In addition, HOCs attended interdisciplinary discharge planning rounds every morning. Their consistent physical presence made the program more visible to hospital physicians and staff. During NPO in the hospital, the HOCs would often be greeted by various hospital physicians and staff in the hallways and HOCs were present at morning rounds. As one hospital participant reported, the HOCs were seen as a part of the team:

When I have the pleasure of engaging with them, then they’re an active part of the patients’ inpatient experience and outpatient experience. They’re definitely a member of the healthcare team. (Hospital_14)

While this was true for General Internal Medicine (GIM), the organizational home for the Navigator Program, for participants who worked on, “every other unit other than GIM, it feels like they’re an external service” (Hospital_13). Participants did not comment on strategies to integrate the Navigator Program onto other services more fully, nor if this would be necessary for success of the program.

While relationships between HOCs and hospital physicians and staff were generally positive and inclusive, participants reported some challenges. These included lack of role clarity, high physician turnover (the hospital is a teaching hospital with residents rotating to a different clinical service each month), and ineffective and limited education about the program, which created confusion among some hospital physicians and staff about the program. In order to support team building within the hospital, the implementation team sent out introductory emails, conducted presentations, and attempted to provide continued education about the program (e.g. the monthly orientation presentation for residents rotating on GIM included a slide on the Navigator Program). Hospital physicians and staff reported that an increase in outreach would have been beneficial for providing clarity about the program and the roles of HOCs.

I don’t know what’s being done. I don’t know what it [The Navigator Program] entails. And so definitely there needs to be some training to the staff, to the Fellows, to all the junior trainees about this. (Hospital_2)

Although the Navigator Program was designed to be a complementary service, participants reported limited clarity in the role of HOCs and function of program. HOCs and the implementation team, in particular, reported that at times hospital staff working on discharge planning would offload their work onto HOCs instead of using them as a complementary service. This could happen because hospital staff are balancing a large patient load and HOCs have specialised training and experience working with people experiencing homelessness. Among the HOCs, there was less confidence that hospital staff would complete tasks for clients. As one participant from the implementation team reflected, the HOCs might feel like, “I know I’m not supposed to do this thing, but if I don’t do it, it’s not going to happen” (Implementation_4). Both points challenged the model component that Navigator Program operated effectively as a complementary program.

The CSPs reported viewing the HOCs as a unique member of a care team, one with increased access to the hospital system broadly, and electronic medical records more specifically. CSPs discussed accepting the HOCs into a team model because they trusted that they had shared values: “I also know how she works and I feel like we have very similar values and so I can also trust that these things will be brought up” (CSP_1). For CSPs, working with the HOCs also made it feel like they had support: “it feels like you are working with a team that’s supporting an individual and you are not trying to reinvent the wheel all the time or do a bunch of detective work to figure out where things are. There is clear idea of what needs to be done and who is going to do it” (CSP_8).

Team-building among the HOCs themselves was something the implementation participants said they had not initially considered, but they reported that it contributed to some of the program’s challenges. This was, in part, because the program started with one HOC, growing as demand increased. As more HOCs were hired, implementers thought that, “as long as the HOCs worked hard and are good at their work then it will be fine” (Implementation_1). However, participants said this approach did not create enough structure to support a team approach to caring for clients, and did not consider who would be responsible for which clients (e.g. shared or individually) and how tasks would get completed. In addition, the need for team work regarding problem solving, brainstorming, sharing resources, and other forms of support (e.g. sick day coverage) had not been considered at the time of implementation. Adjustments were therefore required as the program continued. One clear example was that weekly team meetings became integrated into program operations. The research team joined each week as the HOCs met with the program manager to review clients, provide updates, discharge clients who were stabilized or not engaging with the program, and problem solve as a team. During NPO, the research team saw how important these meetings were for providing stability and support to the HOCs, and to help brainstorm the best ways to support clients. This was also a key aspect of structure that was built into the program to best support the HOCs.

Discussion

This study discusses the implementation of the Navigator Program, a critical time intervention that provides case management and service navigation to people experiencing homelessness during their in-patient hospitalization and for up to several months after hospital discharge. Overall, results indicate that there was a clear need for the Navigator Program, which supported acceptance and uptake of the program. The flexible approach to model design and implementation was essential but a need for more structure became clear once the program grew. Formal and informal approaches to stakeholder buy-in and relationship development with CSPs were successful but at times added strain to the HOCs. An important result was that being a hospital-based program was essential for successful implementation of the program, supporting the HOCs to be integrated into the care teams in the hospital and creating feelings of rootedness.

Obtaining buy-in from all levels of workers and management, both in the inner and outer settings, was key to supporting the implementation of the Navigator Program. A common requirement for buy-in is whether or not those who interact with the program, who make the program feasible, see a need for it [47, 48]. Participants in this study did, suggesting that the standard discharge process did not create an effective bridge from hospital to the community for this patient population, nor did hospital staff have capacity to meet patient’s social needs. Capacity issues are common in the literature, despite more recent advancements in hospitals considering and addressing the social determinants of health [49, 50]. These challenges, which contributed to participants seeing a need for the status quo to change, helped facilitate buy-in from multiple stakeholders.

Developing the program to be simple, cost-efficient, and require little-to-no extra work of hospital staff and physicians was an important design consideration, one that participants suggested led to the program’s successful uptake within the hospital. A study by Gabrielian and colleagues on the implementation of CTI for unhoused veterans in the USA had similar findings, indicating that programs requiring simple changes to existing processes in order to align with the new program were easier to implement [51]. Geerligs and colleagues further confirm that, “The complexity of an intervention can challenge buy-in, and thus developing a simple, easy and low-cost intervention that is flexible to the inner context (e.g. the hospital) are key elements for attaining multi-level stakeholder support.” [52].

An important theme from the results was the benefit of flexibility, but balancing this with structure. The notion of flexibility was discussed in terms of adapting to the hospital context (e.g. inner setting), and supporting an easy referral process. In their systematic review on implementation processes for hospital-based interventions, Geerligs and colleagues found that “interventions that had inbuilt flexibility, and allowed for ongoing change and tailoring, results in greater opportunities to introduce strategies and respond to unforeseen challenges,” [52]. This was also achieved through the slow roll-out of the program. The program was able to adjust as needed, and hospital physicians and staff, along with CSPs were able to see and experience the benefit of the program to their own work and to PEH. Geerligs et al. confirm that, “[T]hose interventions that responded to the hospital context and worked toward ease of integration were more likely to be reported as successful, in terms of adherence, acceptability, and sustainability” [52]. Thus, in practice, a ‘proof is in the pudding’ approach supported the slow development of multi-stakeholder support, and allowed the program to adapt and adjust to the needs of the setting.

The Program struggled with effectively educating and informing hospital staff and physicians, and CSPs about the program. Noted in the results, one issue was turnover of hospital physicians, a consideration especially for academic teaching hospitals where residents rotate frequently. Similar to other studies, this turnover created challenges for sustaining program awareness among this stakeholder group [37, 38]. Outside of the hospital (outer setting), a strategic choice was made to create connections with CSPs informally through their existing relationships with HOCs. Leveraging these existing relationships was an important way to initiate the program uptake; however, it heavily relies on the HOCs to establish and sustain those relationships. It also meant that there wasn’t uniformity in how the program was introduced and CSPs were unaware of the full program offerings. If the success of the program is based on interpersonal relationships and not also supported by more formal processes to develop program recognition, then if an HOC leaves the role and the relationship is lost, it could create issues for program stability. Alternatively, there is a risk that formalization would generate more complex referral pathways or approval from multiple levels of management, thereby jeopardizing the flexibility the Navigator Program–a key implementation success. These are important considerations for future iterations of the program.

Challenges surrounding strategies for educating stakeholders about the program extended to challenges around role clarity of the HOCs. On the one hand, hospital physicians and staff did not necessarily have specialized training or experience working with PEH, and therefore the HOCs fulfilled this specialized role and alleviated some moral distress from hospital physicians and staff (see Cygler et al. [46] for further details on results from analyzing interviews from hospital physicians and staff). This echoes previous research outlined in Table 1 suggesting that a facilitator of implementation is having non-clinical navigators with extensive homelessness-specific experience [32, 36, 38]. On the other, a lack of clarity on what tasks HOCS were meant to do in relation to the current hospital workers created confusion about task allocation. Role clarity has been raised in other studies as an important aspect of program implementation. A study by Goldberg and colleagues of a peer navigator opioid intervention in New York ED departments found that having role clarity and distinction of the intervention deliverers as peers supported the implementation and effectiveness of the program [53]. While that program is different from Navigator in that it uses peers with lived experience as the innovation deliverers (a term in the CFIR), it does show that having a clearly defined role and function supported the ED staff to embrace the program, understand its addition to their own work, and see where it fit into the inner setting. This finding raises considerations if Navigator, or similar programs, should be a complementary (meaning, support with the discharge process for PEH) or a replacement program (meaning, take over leading the discharge process for PEH).

Lastly, designing this program to be embedded within the hospital was arguably the most important implementation decision that has led to the uptake of and support for the Navigator Program. It gave the HOCs access to patient electronic medical records, and office space to work from, and helped to integrate them into the team within the hospital, and among community service providers. These are significant benefits to program implementation. Similar studies have noted that not having access to electronic medical records created challenges for communication and information sharing [32, 36]. Other studies on similar programs supporting PEH from hospital noted that having an office space within the hospital fostered improved relationships and information sharing between the intervention team and the hospital workers [37], and supported easy referrals [54].

Our study has important limitations. It does not directly link implementation data to outcomes data for clients of the Navigator Program. There is a concurrent randomized controlled trial being conducted, and the outcomes data is not yet available. In addition, the qualitative nature of this current study does not allow for determining causation or clear correlation between implementation and outcomes data. In future analyses, however, data from the rest of the process evaluation will be examined to identify potential or possible connections between implementation approaches and experiences in the study as described by Navigator Program clients. Lastly, implementation of programs and interventions is necessarily setting-specific. As such, programs and interventions implemented in other settings need to adapt to their unique contexts while remaining aligned with the core elements of the program or model. Our study findings should be understood with this in mind. This study also has several strengths. We were able to immerse ourselves in the implementation setting, attend numerous meetings about the functioning of the program, and include perspectives of a wide range of stakeholders interacting with the Navigator Program.

Our study is also bolstered by the use of CFIR, a strong framework for exploring program or innovation implementation. Analytically, we found that CFIR at times lacked the specificity we found necessary for the first round of coding and to dig deeper into the data. Using an approach that also allowed for in vivo codes made the analysis more robust. Future methodological research beyond this study might explore quantitative measures of mechanisms supporting implementation, and link those more clearly to program outcomes. Topically, future work that compares implementation of this or similar programs across multiple settings, accounting for the core elements of the program that remain constant and necessary adaptations to each setting, would help provide a more robust model of implementation for scaling up the program. Given that many programs don’t necessarily have public-facing implementation guidance, this would also help create a standard hospital-based case management program model for people experiencing homelessness that other programs could learn and draw from.

Conclusion

A program’s successes or failures can often be connected to its implementation approach, not solely a program’s model. Yet, there is a paucity of research that explores how programs are implemented, what worked, and what did not work. In this study, we highlight the multiple successes in the Navigator Program’s implementation strategy, in addition to key areas that require further improvements in its operations. We discuss the importance of having clear approaches to developing multi-stakeholder buy-in, and the need to allow for a flexible approach to program development that adjusts to the implementation setting, while also creating more systems and structures as the program progresses. Importantly, this study shows that a case management program addressing the health and social needs of PEH may be most successful if embedded within a hospital, allowing for improved team-building among care providers for PEH, which can lead to improved care coordination and, potentially, improved patient outcomes. Overall, the implementation of the Navigator Program holds promising lessons for the development and implementation of future hospital-based CTI case management programs serving PEH.

Data availability

No datasets were generated or analysed during the current study.

References

Hwang SW, Chambers C, Chiu S, Katic M, Kiss A, Redelmeier DA, Levinson W. A comprehensive assessment of health care utilization among homeless adults under a system of universal health insurance. Am J Public Health. 2013;103(Suppl 2):S294-301.

Fazel S, Geddes JR, Kushel M. The health of homeless people in high-income countries: descriptive epidemiology, health consequences, and clinical and policy recommendations. Lancet. 2014;384(9953):1529–40.

Richard L, Carter B, Wu L, Hwang SW. Disparities in all-cause mortality among people experiencing homelessness in toronto, Canada during the COVID-19 pandemic: a cohort study. Front Public Health. 2024;12:1401662.

O’Connell JJ. Premature mortality in homeless populations: A review of the literature. Nashville: National Health Care for the Homeless Council, Inc. 2005.

Funk AM, Greene RN, Dill K, Valvassori P. The impact of homelessness on mortality of individuals living in the United States: a systematic review of the literature. J Health Care Poor Underserved. 2022;33(1):457–77.

Campbell DJ, O’Neill BG, Gibson K, Thurston WE. Primary healthcare needs and barriers to care among Calgary’s homeless populations. BMC Fam Pract. 2015;16: 139.

John W, Law K. Addressing the health needs of the homeless. Br J Community Nurs. 2011;16(3):134–9.

Davies A, Wood LJ. Homeless health care: meeting the challenges of providing primary care. Med J Aust. 2018;209(5):230–4.

Lin WC, Bharel M, Zhang J, O’Connell E, Clark RE. Frequent emergency department visits and hospitalizations among homeless people with medicaid: implications for medicaid expansion. Am J Public Health. 2015;105(Suppl 5):S716-722.

Wadhera RK, Choi E, Shen C, Yeh RW, Joynt Maddox KE. Trends, causes, and outcomes of hospitalizations for homeless individuals: a retrospective cohort study. Med Care. 2019;57(1):21–7.

Zhang L, Norena M, Gadermann A, Hubley A, Russell L, Aubry T, To MJ, Farrell S, Hwang S, Palepu A. Concurrent disorders and health care utilization among homeless and vulnerably housed persons in Canada. J Dual Diagn. 2018;14(1):21–31.

Miyawaki A, Hasegawa K, Figueroa JF, Tsugawa Y. Hospital readmission and emergency department revisits of homeless patients treated at homeless-serving hospitals in the USA: observational study. J Gen Intern Med. 2020;35(9):2560–8.

Canadian Institute for Health Information. Demographic and hospital use statistics for patients experiencing homelessness, 2022–2023. In. Ottawa, ON: CIHI; 2024.

Richard L, Carter B, Nisenbaum R, Liu M, Hwang SW. Identification of homelessness using health administrative data in Ontario, Canada following a national coding mandate: a validation study. J Clin Epidemiol. 2024;172: 111430.

Jenkinson J, Wheeler A, Wong C, Pires LM. Hospital discharge planning for people experiencing homelessness leaving acute care: a neglected issue. Healthcare Policy| Politiques de Santé. 2020;16(1):14–21.

Wang A, Pridham KF, Nisenbaum R, Pedersen C, Brown R, Hwang SW. Factors associated with readmission among general internal medicine patients experiencing homelessness. J Gen Intern Med. 2021;36(7):1944–50. https://doi.org/10.1007/s11606-020-06483-w.

Buccieri K, Oudshoorn A, Frederick T, Schiff R, Abramovich A, Gaetz S, Forchuk C. Hospital discharge planning for Canadians experiencing homelessness. Hous Care Support. 2018;22(1):4–14. https://doi.org/10.1108/HCS-07-2018-0015.

Kushel MB, Gupta R, Gee L, Haas JS. Housing instability and food insecurity as barriers to health care among low-income Americans. J Gen Intern Med. 2006;21(1):71–7.

Wei Lim Y, Andersen R, Leake B, Cunningham W, Gerlberg L. How accessible is medical care for Homless women? Med Care. 2002;40(6):510–20.

Greysen SR, Allen R, Lucas GI, Wang EA, Rosenthal MS. Understanding transitions in care from hospital to homeless shelter: a mixed-methods, community-based participatory approach. J Gen Intern Med. 2012;27(11):1484–91.

Omerov P, Craftman AG, Mattsson E, Klarare A. Homeless persons’ experiences of health- and social care: a systematic integrative review. Health Soc Care Community. 2020;28(1):1–11.

Kertesz SG, Posner MA, O’Connell JJ, Swain S, Mullins AN, Shwartz M, Ash AS. Post-hospital medical respite care and hospital readmission of homeless persons. J Prev Interv Community. 2009;37(2):129–42.

Khatana SAM, Wadhera RK, Choi E, Groeneveld PW, Culhane DP, Kushel M, Kazi DS, Yeh RW, Shen C. Association of homelessness with hospital readmissions-an analysis of three large states. J Gen Intern Med. 2020;35:2576–83. https://doi.org/10.1007/s11606-020-05946-4.

Sadowski LS, Kee R, VanderWeele A, Buchanan TJ. Effect of a housing and case management program on emergency department visits and hospitalizations among chronically ill homeless adults: a randomized control trial. JAMA. 2009;301(17):1771–8.

McCormack RP, Hoffman LF, Wall SP, Goldfrank LR. Resource-limited, collaborative pilot intervention for chronically homeless, alcohol-dependent frequent emergency department users. Am J Public Health. 2013;103(Suppl 2):S221-224.

Tomita A, Herman DB. The impact of critical time intervention in reducing psychiatric rehospitalization after hospital discharge. Psychiatric Serv. 2012;63(9):935–7.

Ponka D, Agbata E, Kendall C, Stergiopoulos V, Mendonca O, Magwood O, Saad A, Larson B, Sun AH, Arya N, et al. The effectiveness of case management interventions for the homeless, vulnerably housed and persons with lived experience: A systematic review. PLoS ONE. 2020;15(4):e0230896.

Reid N, Mason J, Kurdyak P, Nisenbaum R, de Oliveira C, Hwang S, Stergiopoulos V. Evaluating the impact of a critical time intervention adaptation on health care utilization among homeless adults with mental health needs in a large urban center. Can J Psychiatry Revue Canadienne De Psychiatrie. 2022;67(1):57–66.

Lamanna D, Stergiopoulos V, Durbin J, O’Campo P, Poremski D, Tepper J. Promoting continuity of care for homeless adults with unmet health needs: the role of brief interventions. Health Soc Care Community. 2018;26(1):56–64.

Dixon L, Goldberg R, Iannone V, Lucksted A, Bronwn C, Kreyenbuhl J, Fang L, Potts W. Use of a critical time intervention to promote continuity of care after psychiatric inpatient hospitalization. Psychiatr Serv. 2009;60(4):451–8.

Cornes M, Whiteford M, Manthorpe J, Neale J, Byng R, Hewett N, Clark M, Kilmister A, Fuller J, Aldridge R, et al. Improving hospital discharge arrangements for people who are homeless: A realist synthesis of the intermediate care literature. Health Soc Care Community. 2018;26(3):e345–59.

Malden S, Doi L, Ng L, Cuthill F. Reducing hospital readmissions amongst people experiencing homelessness: a mixed-methods evaluation of a multi-disciplinary hospital in-reach programme. BMC Public Health. 2023;23(1):1117.

Manuel JI, Nizza M, Herman DB, Conover S, Esquivel L, Yuan Y, Susser E. Supporting vulnerable people during challenging transitions: A systematic review of critical time intervention. Adm Policy Ment Health. 2023;50(1):100–13.

Stetler CB, Legro MW, Wallace CM, Bowman C, Guihan M, Hagedorn H, Kimmel B, Sharp ND, Smith JL. The role of formative evaluation in implementation research and the QUERI experience. J Gen Intern Med. 2006;21(Suppl 2):S1-8.

Damschroder LJ, Aron DC, Keith RE, irsh K, Alexander SR, Lowery JA. JC: Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science. Implement Sci. 2009;4(50).

Garcia-Jorda D, Fabreau GE, Li QKW, Polachek A, Milaney K, McLane P, McBrien KA. Being a member of a novel transitional case management team for patients with unstable housing: an ethnographic study. BMC Health Serv Res. 2022;22(1):232.

Kahan D, Leszcz M, O’Campo P, Hwang SW, Wasylenki DA, Kurdyak P, Wise Harris D, Gozdzik A, Stergiopoulos V. Integrating care for frequent users of emergency departments: implementation evaluation of a brief multi-organizational intensive case management intervention. BMC Health Serv Res. 2016;16:156.

Forchuk C, Benbow S, Reiss J, Lawson S, Northcott S, Vann R, Catunto D, Jeffrey M, Booth R, Peters A. Preventing discharge to no fixed address– version 2:: evaluation of a best practice program to prevent discharge from hospital into homelessness. International Journal on Homelessness. 2023;3(3):198–213.

Liu M, Pridham KF, Jenkinson J, Nisenbaum R, Richard L, Pedersen C, Brown R, Virani S, Ellerington F, Ranieri A, et al. Navigator programme for hospitalised adults experiencing homelessness: protocol for a pragmatic randomised controlled trial. BMJ Open. 2022;12(12):e065688.

Moore G, Audrey S, Barker M, Bond L, Bonell C, Cooper C, Hardeman W, Moore L, O’Cathain A, Tinati T, et al. Process evaluation in complex public health intervention studies: the need for guidance. J Epidemiol Community Health. 2014;68(2):101–2.

Carmichael C, Smith L, Aldasoro E, Gil Salmerón A, Alhambra-Borrás T, Doñate-Martínez A, Seiler-Ramadas R, Grabovac I. Exploring the application of the navigation model with people experiencing homelessness: a scoping review. J Social Distress Homelessness. 2023;32(2):352–66.

Toronto Shelter and Support Services. Annual Report. City of Toronto; 2023. https://www.toronto.ca/wp-content/uploads/2024/05/97fb-TSSSAnnualReport2023AODA.pdf.

Daily Shelter and Overnight Service Usage. https://www.toronto.ca/city-government/data-research-maps/research-reports/housing-and-homelessness-research-and-reports/shelter-census/.

Damschroder LJ, Reardon CM, Widerquist Opra MA, Lowery J. Conceptualizing outcomes for use with the consolidated framework for implementation research (CFIR): the CFIR outcomes addendum. Implement Sci. 2022;17(1): 7.

Gale NK, Heath G, Cameron E, Rashid S, Redwood S. Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Med Res Methodol. 2013;13(1):117.

Cygler J, Dada OO, Hwang SW, Jenkinson JIR. Evaluation of a hospital-based case management programme for patients who are experiencing homelessness: A qualitative study of health care workers’ experiences with the navigator programme. Can J Gen Intern Med. 2025;16(1):e20240029. https://doi.org/10.3138/cjgim.2024.0029.

Consolidated Framework for Implementation Research. [https://cfirguide.org/].

Kokorelias KM, Shiers-Hanley JE, Rios J, Knoepfli A, Hitzig SL. Factors influencing the implementation of patient navigation programs for adults with complex needs: A scoping review of the literature. Health Serv Insights. 2021;14:11786329211033267.

Raven MC, Carrier ER, Lee J, Billings JC, Marr M, Gourevitch MN. Substance use treatment barriers for patients with frequent hospital admissions. J Subst Abuse Treat. 2010;38(1):22–30.

Gabrielian S, Cordasco KM, Finley EP, Hoffmann LC, Harris T, Calderon RA, Barnard JM, Ganz DA, Olmos-Ochoa TT. Engaging stakeholders to inform National implementation of critical time intervention in a program serving homeless-experienced veterans. Front Psychol. 2022;13:1009467.

Geerligs L, Rankin NM, Shepherd HL, Butow P. Hospital-based interventions: a systematic review of staff-reported barriers and facilitators to implementation processes. Implement Sci. 2018;13(1):36.

Goldberg LA, Chang TE, Freeman R, Welch AE, Jeffers A, Kepler KL, Chambless D, Wittman I, Cowan E, Shelley D, et al. Implementation of a peer-delivered opioid overdose response initiative in new York City emergency departments: insight from multi-stakeholder qualitative interviews. J Subst Use Addict Treat. 2025;168:209542.

Cornes M, Aldridge RW, Biswell E, Byng R, Clark M, Foster G, et al. Improving care transfers for homeless patients after hospital discharge: a realist evaluation. Southampton (UK): NIHR Journals Library (Health Services and Delivery Research, No. 9.17); 2021. https://www.ncbi.nlm.nih.gov/books/NBK574259/. https://doi.org/10.3310/hsdr09170.

Lukas CV, Holmes SK, Cohen AB, Restuccia J, Cramer IE, Shwartz M, Charns MP. Transformational change in health care systems: an organizational model. Health Care Manage Rev. 2007;32(4):309–20.