Case Studies: ART and PrEP Implementation Approaches for People Living with Homelessness and Housing Insecurity

In this section, we ground our theoretical framing of determinants and intervention functions to address PrEP use with recent case studies of oral and long-acting ART and PrEP implementation approaches for people living with homelessness and housing insecurity in the United States. These case studies offer a data-driven approach to identify common components and their effects on HIV outcomes (Table 2). Findings presented here reflect published data but most of these programs are ongoing and evolving, and their findings and package of services may differ at the time of this manuscript publication.

Table 2 HIV treatment and prevention programs offering services to people experiencing homelessness or housing insecurityOral ART Delivery for People Living with HIV

Three programs, POP-UP [12, 33], the MAX Clinic [13, 34], and HHOME [35] focus on low-barrier, high-touch care delivery to promote ART uptake and adherence among people living with HIV who are unstably housed. The POP-UP program is housed within Ward 86, the safety-net HIV clinic at San Francisco General Hospital, and operates through joint efforts from the San Francisco Department of Public Health (SFDPH) and the University of California. It provides low-barrier HIV primary care for unstably housed people who are viremic and have difficulty engaging in scheduled appointments [12, 33]. The program includes four patient-facing ART implementation approaches: low-barrier comprehensive care with integrated substance use and mental health care services; social work and case management including referrals for housing assistance; financial incentives; and enhanced outreach and care coordination [12, 33]. In addition, POP-UP includes provider-facing implementation approaches to facilitate patient tracking and care coordination including clinical trainings, weekly case conferences, data dashboards, and quarterly team support sessions [12, 33]. The program has enrolled a high-need cohort (e.g., 70% of patients had a substance use or mental health disorder) and reported improvements in clinic visit attendance and HIV viral suppression over a 12-month period compared with baseline [12, 33]. The MAX Clinic is housed within the Harborview Medical Center and is operated by the Madison Clinic and Public Health – Seattle & King County. It enrolls a similar cohort of patients with high viral loads and challenges attending clinic visits in Seattle, WA and also includes the same four patient-facing implementation components as in POP-UP along with three provider-facing components (low case loads with the assistance of a cadre of medical case managers, automated staff alerts for patient hospital admissions, and coordination to plan care transitions) [13, 34]. In a matched pairs analysis, enrollment in the MAX Clinic significantly improved viral suppression over a one-year period [13, 34]. The HIV Homeless-Health Outreach Mobile Engagement (HHOME) program is operated by the San Francisco Department of Public Health (SFDPH). It offers mobile care, with a mix of street- and clinic-based service delivery, to people who are experiencing homelessness or unstable housing and have detectable HIV viral loads, active substance use disorders or severe mental illness, and high emergency department or hospital utilization in San Francisco [35]. HHOME also includes mobile outreach-based primary care (with care in shelters, streets, encampments, hospitals, and treatment programs as well as at an HIV drop-in clinic), integrated substance use treatment and housing case management, but does not provide financial incentives [35]. Providers also attend weekly case conferences and engage in intensive team communication to coordinate patient care [35]. From 2014–2017, the majority of HHOME clients (79%) achieved HIV viral suppression in a one year period and 84% transitioned to stable housing. [35]

Status-Neutral ART and PrEP Delivery

Two programs, University of Alabama at Birmingham’s “Ending the HIV Epidemic: Addressing HIV Health and Homelessness” (AH3) [36] and the SFDPH’s Whole Person Integrated Care (WPIC) models [14], provide status-neutral HIV testing and linkage to oral and long-acting HIV treatment and prevention services (Table 2). Of the case studies described here, AH3 is unique in offering HIV services within homeless shelters by providing integrated case management, HIV counseling, and testing to people participating in daytime activities or staying overnight at a shelter in Alabama [36]. The program does not describe other patient- or provider-facing implementation approaches; however it does describe general operations. Over a one-year period, recent reports show that all shelter guests accepted HIV testing and were counseled on PrEP however only 40% of those who tested positive for HIV attended a clinic appointment within one month and only 1.4% of those who tested negative for HIV expressed interest in PrEP. [36]

The SFDPH WPIC program provides low-barrier oral and long-acting injectable ART and PrEP for people experiencing homelessness or unstable housing in San Francisco [14]. Patient-facing program components include “direct-to-inject” long-acting ART and PrEP (without the requirement for an oral medication lead in) integrated within primary care through a low-barrier (e.g., no appointments) central open-access clinic, satellite shelter health clinics, mobile street-based outreach, and partnerships with local community-based programs providing housing assistance [14]. The program also offers incentives for on-time injections and laboratory draws and enhanced monitoring and care coordination via multidisciplinary clinic- and street-based care providers [14]. Between 2021–2024, 22 clients initiated on long-acting ART (15 with detectable viremia) and all but 1 achieved and maintained viral suppression. In addition, 19 clients initiated long-acting PrEP, all of whom remained HIV negative [14]. Fewer than 10% of long-acting ART and PrEP injections were delayed overall.

Oral and Long-Acting PrEP Delivery

Two programs, Boston Health Care for the Homeless Program (BHCHP) [15, 16] and the Ward 86 clinic [37], developed programs focused on providing low-barrier PrEP to people experiencing homelessness or unstable housing (Table 2). The BHCHP program focuses on oral PrEP delivery for PEH engaging in risky sexual behavior or drug use and provides PrEP education and low-barrier PrEP integrated with primary care services, on-site medication storage, short-term refills, and intensive outreach and navigation support [15, 16]. It also explicitly emphasizes developing a culture of trusting and respectful patient-provider relationships to promote PrEP uptake and adherence [15, 16]. From 2018–2020, the program provided twice the number of PrEP prescriptions than in 2017 and, of 152 clients who received PrEP prescriptions, the cumulative probability of PrEP persistence through 6 months was 44%. [15, 16]

The Ward 86 program offers oral and long-acting PrEP to a patient population that includes those experiencing homelessness, mental health challenges, and substance use disorders via a “direct-to-inject” approach and use of drop-in appointments coupled with incentives and in-clinic provision of harm reduction supplies, along with a weekly multi-disciplinary panel management meeting [37]. It includes but is not limited to patients from the POPUP program, which is described above. As recently described, the program initiated 30 patients on long-acting PrEP in a 15-month period (from March 2022 to June 2023) and, similar to findings from the SF WPIC program, the majority of injections were on time and all patients remained HIV negative through follow-up. [37]

The SHELTER Multicomponent Implementation Strategy for PrEP Provision among People Experiencing Homelessness and Unstable Housing

Based on the behavior change targets theorized to promote PrEP use among people experiencing homelessness and unstable housing and case studies of ART and PrEP delivery models for this population, we propose the “SHELTER” (Supporting HIV prevention through Enhanced, Low-barrier, high-Touch services and Empowering Resources) multicomponent implementation strategy (Fig. 1). SHELTER comprises a minimum set of six core components to advance oral and long-acting PrEP delivery, identified by combining components found in most or all of the case studies that also meet criteria for candidate intervention functions identified in our BCW framework (Table 1). SHELTER components are applicable to all PrEP options: daily oral PrEP, event-driven oral PrEP, and injectable and long-acting PrEP options (which may be uniquely beneficial for unstably housed people given adherence challenges in this population).

Fig. 1

The “SHELTER” multi-component implementation strategy to promote oral and long-acting PrEP delivery among people experiencing homelessness or housing insecurity

In Table 3, we specify the individual elements of the SHELTER implementation strategy using a unified framework and established guidelines from the implementation science field for naming, defining, and operationalizing each to ensure that they are well understood and enable reproducibility [38]. Individual components of the SHELTER strategy have been derived from those described across case studies. Outreach, care coordination and street-based care coupled with low-barrier clinic spaces, are critical to reach individuals experiencing homelessness or housing insecurity and provide care. While traditional clinic-based care is generally more thorough, has more capacity, and can more effectively link patients to ancillary care, newer models of care delivery show that street-based care is an additional option particularly well-suited for people experiencing housing instability. Low-barrier comprehensive care and financial incentives for visits, laboratory testing, and PrEP adherence are critical to meet patients’ needs and promote continued engagement, while multidisciplinary provider meetings and data tracking are necessary to facilitate care coordination. Case management (with case managers based in clinics and in partner organizations in communities) can further assist patients in reaching PrEP goals and in linking to additional services to support with housing, substance use and mental health treatment, and trauma recovery. Underlying these components is the cultivation of a trusting relationship between the healthcare team and PrEP clients. By branding these components together into the SHELTER strategy, we hope to promote broader implementation of a comprehensive low-barrier, high-touch PrEP delivery approach that can be evaluated as a package and facilitate comparison of findings on PrEP outcomes across different settings and component operationalization approaches. We also provide suggestions of key implementation, service, and/or client-level outcomes that could be measured for each SHELTER strategy component, based on what has been measured in case studies including these components and the Proctor Implementation Outcomes Framework (“IOF”, which theorizes that implementation outcomes could lead to service outcomes which could in turn lead to client-level outcomes) [39]. These are meant to be starting points for evaluation planning evaluations of SHELTER components but we do not offer descriptions of how to operationalize each IOF outcome for each component.

Table 3 Specifying components [38] of the SHELTER implementation strategy to promote oral and long-acting PrEP delivery for people experiencing homelessness or housing insecurity

Low-barrier comprehensive care includes any model of flexible oral or long-acting PrEP delivery alongside other services to address barriers to PrEP in this population (e.g., harm reduction or ambulatory care services, drop-in clinic hours, telehealth visits, same-day starts, direct-to-inject long-acting PrEP, onsite options for laboratory testing and medication storage, and linkage to trauma recovery, mental health care and/or substance use treatment). However, it does not necessarily include full scope primary care which would be difficult and cost- and logistics-prohibitive to offer in many settings. Based on theoretical underpinnings of the BCW framework,[40] we hypothesize that the action target for this component is altering or reducing social and environmental barriers to PrEP, providing tangible resources for clients, and altering organizational culture and climate at the clinic level. Both status-neutral clinics (e.g., WPIC) and HIV-specific clinics (Ward 86) have PrEP programs, though primary care services are more comprehensively offered in the former. Moving forward, it would be ideal to equip both types of clinics for PrEP delivery using this comprehensive low-barrier care approach to address co-occurring housing, mental health, and substance use needs. Evaluating success of this component could include measuring changes in implementation outcomes like PrEP acceptability, and appropriateness of PrEP and other health care service delivery by clients and service outcomes like PrEP uptake and adherence. Secondary measures of success could include reductions in risky behavior, reductions in substance use, improvements in mental illness, and reductions in medical mistrust. Among providers, key implementation outcomes could include acceptability, feasibility, and sustainability of the model.

Integrated case management is defined as any care coordination for PrEP and/or other health or non-health (e.g., housing assistance) related services, which could include referrals and warm hand-offs for trauma recovery, mental health or substance use care. This component is hypothesized to target client resources, action planning around PrEP use and broader health and wellbeing goals, and self-efficacy to access resources. At the clinic level, it also could change organizational culture and climate around integrated service delivery and professional roles. Key implementation outcomes that might be affected by this component include acceptability of health care services broadly, penetration of PrEP (i.e. uptake of PrEP by those who could benefit from it), trauma, mental health, substance use, and housing assistance services, and fidelity to referral protocols.

Incentives include any reward contingent on health seeking behavior related to PrEP or care engagement. They are inconsistently included across our featured ART and PrEP delivery case studies, which may be due to limited resources. However, they are included here based on their effectiveness in PrEP and non-PrEP environments that include them, as well as their theoretical potential. Specifically, incentives reward client behavior; they theoretically motivate and empower PrEP clients to maintain engagement despite competing priorities related to other health risks and food, financial, and housing insecurity. The value of formal contingency management for PrEP adherence has been studied in multiple contexts,[41, 42] but PrEP programs will need to consider whether to and how best to structure an incentive program based on needs of clients who are experiencing housing insecurity.

Enhanced outreach and care coordination are defined as any navigation services to reach PrEP clients outside of clinic settings via, for example, street-based outreach teams or patient navigators. This component could include PrEP awareness and navigation in homeless shelters, tent encampments, and other community settings and provision of off-site injections or reminders regarding upcoming clinic-based injections to patients via phone, telehealth, or mobile street-based outreach. By increasing community knowledge about PrEP and improving communication pathways, this component could improve social norms around PrEP use. It also has the potential to reduce barriers to PrEP initiation and refills through access to PrEP navigators who can troubleshoot challenges outside of clinics. Potential implementation outcomes of this component include PrEP acceptability and PrEP penetration in communities of individuals who are homeless or unstably housed. Potential service outcomes include changes in PrEP uptake and adherence and durations of periods off PrEP.

The proposed SHELTER model includes two provider- or clinic-facing components: provider team communication and data tracking. Provider team communication includes any communication (e.g., in-person, phone-based, contact through the electronic health record) to coordinate patient management across multidisciplinary teams of health workers. Team communication strategies should attend to coordination needs within and between care delivery systems. Most case studies operationalize this as weekly case conferences to discuss individualized care plans, along with regular email and phone-based communication across providers. Regular and clear communication streams improve clinical knowledge about clients’ needs, skills to engage with the client across the provider team, and goals and action planning for the client’s PrEP use. Data tracking includes any data dashboards or automated alerts to monitor patient clinic visit attendance, PrEP use, or health outcomes. The MAX Clinic, POP-UP, and WPIC programs also use automated alerts (e.g., via Epic dashboards) to indicate when patients are seen in the emergency room or admitted to the hospital [13, 34], although these alerts may be difficult to implement at scale depending on their volume. Data dashboards and alerts could assist the provider team in decision-making around PrEP prescribing and other care and monitoring client’s care engagement. Key implementation outcomes of both the provider team communication and data tracking components include potential improvements in care engagement for services, improvements in care retention, fidelity to PrEP delivery protocols and increased acceptability, feasibility, and sustainability of PrEP delivery for people experiencing homelessness and unstable housing among clinical team members.

Underpinning all SHELTER components is the need to cultivate and maintain strong relationships of trust between PrEP clients and providers and between provider teams and community partners. For example, PrEP providers with BHCHP described the importance of trusting relationships with their clients in promoting safe spaces for conversations about PrEP clients’ broader needs and challenges [15]. While other case studies did not explicitly describe how they developed or maintained trusting relationships between provider teams and patients, it was clear that all created long-standing partnerships with community stakeholder groups and had deep ties with local organizations also providing substance use, trauma, mental health, and housing assistance to promote integrated service delivery. The importance of patient-provider relationships has also been found in quantitative research on preferences for HIV service delivery among PEH, who would trade financial gain for personal relationships with their care team [43]. By including trusting patient-provider relationships in our SHELTER model, we hope to encourage PrEP providers to recognize them as a necessary criteria for PrEP delivery among harder to reach populations and to specify ways that they can cultivate these relationships and measure their success over time in research and dissemination products.

Future Directions

Our SHELTER multi-component implementation strategy offers a specified and comprehensive approach to oral and long-acting PrEP delivery programs for people experiencing homelessness and unstable housing but is based on a limited evidence base of case studies from the HIV prevention and treatment space. This review identifies a number of gaps and future research directions that are needed to maximize the utility and effectiveness of the SHELTER strategy on PrEP uptake and adherence in this population. The majority of case studies included here based their findings off pre-post test analyses and none disentangled the effectiveness or cost-effectiveness of specific ART and PrEP delivery components on uptake and adherence. There remains a need for consideration of adaptations and evaluations of the overall SHELTER strategy and its individual components, particularly given the time- and resource-intensiveness of this strategy. This review also has a number of limitations related to generalizability that need to be considered. First, this research literature and the included case studies primarily come from Democrat-led West Coast states with Medicaid expansion and funding for wrap-around health care services. While one case study was from a site in the southeastern United States, other information is not available on how components of SHELTER could be feasibly, acceptably, and effectively delivered in places with less funding and the highest HIV burden. Second, programs like Ryan White and Housing Opportunities for Persons with AIDS (HOPWA) provide funding for wrap-around healthcare services for people living with HIV but gaps remain in understanding how to offer status-neutral services for individuals eligible for and seeking HIV prevention medication alongside other housing and healthcare. Third, we were limited to reporting information from the published literature and conference abstracts, but there may be other programs offering PrEP in innovative, multi-component packages for people experiencing homelessness and unstable housing that are not published in peer-reviewed databases. Future work is needed to understand how wrap-around services are offered in different regions of the United States with different funding climates for HIV prevention, housing, mental health, and substance use service delivery to explore where SHELTER may be most applicable.

Evidence is mixed on the effectiveness of financial incentives on improving medication adherence, and it is unclear through which behavioral or cognitive pathways these incentives may work and what their ideal timing or amounts should be [18]. We include incentives in SHELTER but they may not be appropriate for all PrEP programs. Contingency management for PrEP adherence (e.g., using biomarkers of tenofovir adherence or other objective metrics) may also be empowering. Finally, all intervention functions identified using our COM-B framing are encompassed by one or more SHELTER components except for peer modeling around PrEP use. In populations with low PrEP awareness and high perceived or experienced PrEP stigma and medical mistrust, peers can be a welcoming and trusted source of PrEP information [30]. One recent pilot study conducted with US women offered PrEP education, counseling, and referral via peers at mobile syringe exchange sites and sex worker and syringe exchange drop-in centers [44]. Approximately 40% of this population reported housing insecurity at enrollment and, while PrEP interest was high (73%) after receipt of this peer intervention, only 6% attended an initial PrEP appointment and 0% received a PrEP prescription [44]. However, streamlined PrEP delivery approaches with peer navigators who could offer same-day PrEP services may mitigate some challenges with this peer approach.

We are at an exciting point of increased availability of injectable PrEP that can be administered once every two months in the United States, with other long-acting injectable and oral options on the horizon. These long-acting modalities can address many of the PrEP barriers related to capability (e.g., difficulty remembering to dose) and opportunity (e.g., lack of places to store medication) described in Table 1. However, long-acting PrEP is not a panacea to address intersectional determinants of health, including substance use, trauma, and mental health challenges. While the frequency of use and operationalization vary for oral versus long-acting PrEP, all SHELTER components apply for all current PrEP modalities with regular client contact and case management remaining necessary as long-acting PrEP is scaled up. Case studies also described the importance of same-day PrEP starts to reduce drop-offs between referrals, laboratory testing, and PrEP initiation. Long-acting PrEP programs offering same-day start have been found to be both safe and effective,[14, 37] although challenges exist with having clinic-based stock supply available and continued efforts to streamline long-acting PrEP initiation are critically needed among people experiencing homelessness and housing instability.

The SHELTER strategy focuses on implementation approaches to advance PrEP delivery in health care and community spaces by reducing barriers related to appointment constraints, wait times, and referrals for additional services among people who are experiencing homelessness and unstable housing and offering PrEP outreach, medication delivery and injections, laboratory draws, and patient navigator visits in trusted community locations. Several case studies (e.g., HHOME, WPIC, BHCHP) considered approaches to move PrEP into community-based locations like those also providing substance use interventions or homeless shelters. As one model of this, PrEP programs could leverage community partnerships to offer “prevention on demand”, where individuals signed up for PrEP through a clinic or mobile outreach could pick up medications at a local pharmacy or other community location of their choice. Two pilot studies offered PrEP information, counseling, and referrals at community locations, although they had very low PrEP uptake [36, 44] indicating that more work is needed to translate knowledge about PrEP to PrEP uptake among individuals experiencing homelessness and unstable housing. Co-locating PrEP providers, who can offer same-day pills and injections, at venues where these individuals already are (e.g., substance use treatment programs, syringe exchange programs, soup kitchens, homeless shelters), remains critical. As an example of this the WPIC program in San Francisco offers direct-to-inject long-acting PrEP in shelters, tents, and other community-based locations [45]. The outreach, care coordination, case management, and provider tracking and communication aspects of the SHELTER strategy can apply to these de-medicalized contexts to offer low-barrier comprehensive care with PrEP outside of health clinic settings.