Prioritizing Patient Perspectives When Designing Intervention Studies for Homeless Older Adults

Abstract

Purpose:

Medical respite provides postacute care to people experiencing homelessness upon hospital discharge if they are too sick to recover on the streets or in a traditional shelter. The current study examined the feasibility of conducting a study to test the effectiveness of a medical respite intervention for older people experiencing homelessness.

Methods:

Fifteen patient and 11 provider participants were interviewed between July and November 2018.

Results:

Participants’ considerations for how to design a program of research included (1) desired qualities of researchers; (2) preferences for study design; (3) mechanisms for participant recruitment and retention; (4) what, where, and how to collect data; and (5) barriers and motivations to participation.

Conclusions:

Findings from this study build on an emerging research base on how to appropriately engage vulnerable patient groups, including older people experiencing homelessness, in trauma-informed research by including peer researchers on research teams to serve as advisors throughout the research process.

Keywords

homeless hospital hospital discharge medical respite

With the goal of providing safe locations for homeless persons to continue medical recovery following hospital discharge, medical respite programs have improved health for general populations of homeless people and reduced health care utilization and costs (Doran et al., 2013). In North America, medical respite describes the provision of postacute medical care for homeless people who are not ill enough to warrant staying in a hospital bed yet are too sick or frail to recover from a physical illness or injury on the streets or in a traditional shelter (Doran et al., 2013; National Healthcare for the Homeless Council, 2012). These programs have also been called shelter-based convalescence, “hoptel,” intermediate care, and recuperative care (Cornes et al., 2018; Doran et al., 2013). The first medical respite programs were developed in the mid-1980s in the United States and, today, more than 200 Health Care for the Homeless programs operate nationally and are integrated into the U.S. federal health care system (Zlotnick et al., 2013). In Canada, the first medical respite program was not developed until 1999, and the few formally recognized programs are in Ontario (National Health Care for the Homeless Council, 2013). In British Columbia, there are a few rehabilitation beds available in shelter or transitional housing to individuals who are experiencing homelessness and require ongoing care/support (Canham, 2019).

As an intervention, the two basic models of medical respite are shelter-based models and freestanding respite units (Ciambrone & Edgington, 2009). Within these two models, the range and scope of services vary, though typical services include medical care from a nurse, physician, physician assistant, and/or nurse practitioner; medication management (storage, dispensing, and sometimes pharmacy); case management (income assistance, benefits acquisition, housing placement, health education, etc.); and onsite or referral to mental health/substance use treatment, transportation, food, and safety/security (Ciambrone & Edgington, 2009). In addition, there is variation in patient eligibility and length of stays (Bauer et al., 2012; Dorney-Smith et al., 2016; van Laere et al., 2009; Zerger et al., 2009). However, most programs require patients to be adults in need of medical treatment who are homeless, lack family support, and not actively using drugs or alcohol; some programs require patients to have the capacity for self-care, while others can support patients with limitations in the management of their activities of daily living (Buchanan et al., 2006; Buchanan et al., 2009; Doran et al., 2013; Sadowski & Buchanan, 2009; Zerger, 2006).

Despite the recognition of the unique health care needs of older people experiencing homelessness (OPEH; Brown et al., 2017; Canham et al., 2020), defined herein as age 50+ years (C. I. Cohen, 1999), limited research has explored medical respite design considerations to match the needs of OPEH (Canham et al., in press), and no research has examined the effectiveness of medical respite for OPEH. While Doran et al. (2013) recommend the need for a randomized controlled trial (RCT) study to test the effectiveness of medical respite programs for general populations of people experiencing homelessness, the feasibility of conducting a medical respite intervention and the unique considerations for OPEH remain unknown. As a first step in designing a study to test the effectiveness of medical respite for improving health outcomes for OPEH and reducing health care costs, we examined three key components of feasibility (e.g., the process and logistics; Eldridge et al., 2016): study design, access to patient participants, and data collection procedures.

Method

At the study outset, a steering committee, which included the first author, three lived experience partners, and 14 multidisciplinary professionals in the health and social service sectors with experience coordinating hospital discharges of OPEH, was formed. The committee was consulted quarterly as key decision makers over the yearlong study.

Participants

Patient and provider participants were recruited with the assistance of the steering committee, who invited their colleagues and clients to contact the first author if they were interested in the study. Criteria for study participation included being age 19 or older; involved in the direct delivery or receipt of health or social services for OPEH; and able to speak conversational English, participate in an interview for up to 1 hr, and give voluntary and informed consent. We did not limit patient and provider participant age criteria to age 50+ to ensure we could capture the perspectives of participants of all ages who have an opinion on how to design a medical respite intervention study for OPEH. Written informed consent was obtained from all participants. Ethics approval was obtained from a university institutional review board, and participant names were removed to protect identities.

Fifteen patient participants who self-identified as male (n = 10) or female (n = 5) and aged 36–73 years (M: 54 years) were interviewed; five reported living in social housing, six in a transitional shelter, three in a general shelter, and one reported precarious housing at the time of the interview. Most patient participants self-reported race/ethnicity as Caucasian (n = 10), while four identified as Indigenous, one identified as Asian, and one identified as Black. The 11 provider participants (five male, six female) were professionals working in health care, shelter, or housing sectors across Metro Vancouver. Five provider participants were social workers or nurses employed by a health authority who supports the discharge of people with no fixed address from the hospital, three were staff working in shelters, two had an academic affiliation and clinical responsibilities (internal medicine and psychiatry), and one was a housing services worker for a social service agency. No additional demographic data were collected from provider participants.

Data Collection

One-on-one or focus group interviews were conducted according to the preferences and format most appropriate and acceptable to patient and provider participants. All participants indicated their preference for an in-person or phone interview or focus group, and all preferences were accommodated. Between July and November 2018, 11 patient participants were interviewed by phone (n = 3) or in person (n = 8), and four participated in an in-person focus group. Between September 2018 and February 2019, nine provider participants were interviewed one-on-one by phone, and two participated in an in-person focus group. Patient participants received a CA$25 gift card for their participation; no provider participants were compensated.

A semistructured interview agenda (Appendix A) was developed and approved by the steering committee to include modified questions from the existing literature to fit the local context and others that related to the study objectives. Example questions included: How can we most successfully recruit older patient participants? What would affect an older patient’s willingness to participate in this research? What data should we propose to collect for the next stage of research? What barriers would there be to the collection of data? The interview guide developed for patient participants slightly differed from that developed for provider participants in order to capture their distinct viewpoints, but the questions remained the same regardless of interview format. In-depth, detailed notes were taken for five patient participants who chose not to be audio recorded or whose interview was conducted in a noisy setting (e.g., coffee shops). All other interviews were audio recorded and transcribed verbatim.

Data Analysis

Using inductive analysis to allow themes to emerge from the data, five phases of thematic analysis (Braun & Clarke, 2006) were conducted. The first author began by reading and rereading the transcripts to gain familiarity with the data. This author subsequently generated an initial set of descriptive codes closely related to the text (Boyatzis, 1998) by identifying patterns of meaning present in the data. Next, all codes generated in Step 2 were collated and organized into an initial set of themes, which was shared with the steering committee for feedback. Following this review, the list of emergent themes was then further refined by the first author by removing, separating, and collapsing them. Finally, the set of themes reported below were named and defined and once more reassessed by the steering committee. Through each step, NVivo Version 12 qualitative data software (QSR International, 2018) facilitated data management and analysis.

Results

Participants described five considerations for how to design a program of research to examine the effectiveness of a medical respite intervention for people with lived experience of homelessness, including (1) desired qualities of researchers; (2) preferences for study design; (3) mechanisms for participant recruitment and retention; (4) what, where, and how to collect data; and (5) barriers and motivations to participation.

Desired Qualities of Researchers

Participants described several qualities that researchers should have for a program of research on medical respite to be successful. First, they underlined qualities that would aide participant recruitment, retention, and data collection. Participants emphasized that researchers need to build trust with potential research participants. As one patient participant (PA5) stated, “Trust is a privilege.” According to a provider participant (PR5), the research team needs to demonstrate to research participants that they are listened to and valued. This is particularly important for homeless patients who may mistrust the health care system and persons in positions of power: “They’ve started to see that nobody is listening to them, nobody is doing anything, then that starts to compound on them as well. And then it starts that mistrust of service providers….”

Participants also emphasized that researchers be flexible, patient, and nonjudgmental when conducting research in a medical setting because of the unease that some people may experience when involved with institutionalized care.

You’d have to be very flexible because everyone has a different kind of reaction when they’re in a health care setting…So, just being flexible in that, and knowing that it’s not a comfortable environment for everybody, and so some people might require a little more time and rapport building and flexibility…Indigenous people, in particular, it could be very challenging, given the history of everything that’s happened. (PR11)

Patient participants, in particular, highlighted that empathy and contextual awareness were important because of the historical trauma and mental pain that often accompany potential patient participants’ physical health. One patient participant (PA15) noted that researchers should “probably [have] some level of emotional connection with the trauma that’s going on…” Another (PA4) agreed:

It’s just taking a really skilled person who has a lot of empathy and understands physical illness or injury. Is it going to trigger and bring up other pain that is not physical for that person? Because that’s kind of what [the] hospital is about.

To overcome the challenges of mistrust, and to facilitate empathy and understanding, participants agreed on the value of involving peer researchers. A peer researcher was described as someone who would be able to relate to potential participants on a more intimate and equal level because they have also experienced homelessness.

Well, a peer research assistant would be someone who’s a peer who either is homeless or has been homeless and they’re older and they understand what they’re looking for. Because they can reach out to these people so much easier than someone walking down the street…. (PA10)

Peer researchers were also identified as better positioned to create a sense of safety among potential participants and to obtain more honest reports.

I firmly believe in peer-led, peer-based conversations. So, I’ve been a part of research groups and proposals where there’s actually a peer-designated researcher role within the research team to reach out and to talk to people and so there is that level of acceptance and understanding of community, and that’s really a level playing field for trying to get those real conversations and a sense of safety. (PR10)

Preferences for Study Design

Acceptability of an RCT

When asked whether it would be acceptable to have a treatment and control group that, respectively, does and does not receive medical respite services, patient participants suggested that this would be no different than their current experience. One patient participant (PA3) stated, “I guess it would be sort of business as usual, right? But you would feel slighted somewhat, for sure. But I think that if you’re being part of a study, there can’t be any expectations.” While randomization of potential participants into treatment and control arms was not considered unethical by providers, one (PR11) suggested that if study participants knew there were medical respite services that they were unable to access, this could be “discouraging.” Nevertheless, an RCT was also understood as an important first step toward advocating for increased respite beds. One provider (PR5) stated,

If you’ve got a certain amount of funding to only have a certain amount of beds open and to run it like that—I mean, it’s unfortunate, but that’s the way it is. I guess what we’d have to see is whether or not that model works….

Some provider participants did suggest an alternative to an RCT design to avoid creating another experience of inequity for participants. Instead, they suggested conducting a program evaluation of an existing medical respite program, inclusive of a cost-effectiveness analysis. A program evaluation study could look at processes and outcomes of the intervention and take an individual- or systems-level approach. At the systems level, provider participants highlighted the importance of examining how a medical respite program would impact acute care length of stay and discharge planning and the allocation of resources. One provider (PR10) indicated that there is value in “using a population health data approach to put some monetary value on [medical respite].”

Qualitative, quantitative, or mixed methods?

Participants had diverse perspectives on the methods a future study should use. Patient participants, in particular, indicated a preference for open-ended, qualitative methods over quantitative designs because the latter often restricted choices without capturing someone’s full experience. One patient participant (PA2) stated,

Speaking face-to-face, or talking like this, is more advantageous than multiple choice. Because multiple choice, you pick one, whether it’s right or not. And if you’re speaking face-to-face then you’re voicing your opinion and no one is prompting you in one direction or another.

In-person data collection was suggested by another patient (PA8) to promote more honest reporting:

By interviewing them, that’s the only way. If you phone me, I’m going to lie to you. If I’m face-to-face, you know what’s lies and what’s not. So, if you interview them [do it] face-to-face….

One patient participant (PA11) suggested using mixed methods:

How about doing a survey where you have some questionnaires on a document, you have a group of us and they fill out some of the questionnaires, some of the responses on this document, and then you communicate about them afterwards?

A provider participant (PR2) agreed that mixed methods would lead to “rich” findings:

I do think some flavor of the patient experience would be great. And it may not be in the standardized measure, it may be more qualitative. It would be selecting people, and…what was successful for them with respite, what it allowed them to do—versus their previous experiences being discharged from hospital.

Mechanisms for Participant Recruitment and Retention

Participants overwhelmingly identified the hospital as a primary location for recruitment into a program of research on medical respite and hospital social workers as key to recruitment. A patient participant (PA11) stated, “I would say talk to social workers at the hospital, give them the information, and have them pass the information on to the clients before they leave.” One provider suggested that potential participants can be identified through hospital codes that indicate which patients do not have a discharge location.

Some participants suggested that an ideal time to engage potential patient participants was immediately following discharge from hospital. The immediacy of the experience would highlight the importance of the research and enhance their willingness to engage in it.

For a lot of people, especially in this walk of life, the immediacy of it is probably important. They seem to have a lot more to say, even to us [providers], when they do just get out [of the hospital], rather than when you approach a week later or 2 weeks later. (PR7)

When one patient participant (PA14) was asked how willing they would be to be approached about a research study when being discharged from the hospital, they stated, “That would [have] probably be[en] the perfect time.”

On the other hand, researchers need to remain mindful that post-discharge can be a challenging time, particularly for marginalized populations. A patient participant (PA4) suggested that potential participants be recruited just before discharge and then approached again later to complete the research activities after they have had time to adjust to their discharge:

When someone is coming out of the hospital setting, that’s probably, obviously, not the best time to approach someone…. Sometimes there’s a lot of pain involved. Just the agony of trying to deal with those immediate needs, for me is a little challenging. I knew about this research prior, so if there were some way to recruit people prior to being discharged, or just as they’re coming out, that might be a little easier because we’re not in that zone yet.

Other recruitment sources identified by participants included detox programs, rehabilitation centres that care for high numbers of OPEH, outpatient clinics, community health nurses, community case managers who work for shelter/housing organizations, and outreach workers. Finally, without using the term “snowball sampling,” participants indicated that their peers are an important source of recruitment.

Participant retention

Participants suggested that a challenge to participant retention in longitudinal research is unstable housing and limited access to online resources or phones. To stay engaged with OPEH over time, participants suggested that study participants be attached to a case management team or caseworker who researchers have permission to contact if they are unable to locate the study participant. In addition, it was suggested that multiple forms of contact information be collected from study participants and be regularly updated (e.g., telephone numbers, email addresses). As well, making note of the common locations frequented by study participants can assist in locating study participants. A provider (PR6) stated, “Catching them in areas where they might be and whether there’s drop-ins where they could flag those folks. The shelters certainly do have a good eye for this because they know these guys need a ton of support.”

Incentives for participation

According to one provider participant (PR1), financial incentives are important considerations for participant retention because “people that are in that situation don’t have a lot of money and that’s always a struggle.” Another provider (PR10) similarly suggested:

To be very honest, any sort of incentive-based participation always works well, just because the world of medical respite that I’m sitting in—it is with a very marginalized population. So, there has been a lot of success with some of the programs that we’ve run where there’s gift cards for time, and any sort of food security.

While this provider felt gift cards might be more appropriate than cash so that research participants were restricted in how they could spend the honorarium (e.g., not on alcohol or drugs), participants generally agreed that cash honorariums should be offered as compensation.

People [who] work on the front line in the community, they get roughly CA$20 an hour. So, it’s kind of keeping them in line with that, and just gives them a bit of incentive. I think it’s really important that we value that their time is also important. So [whether] it’s an incentive or an honorarium or however we can do it, I think it’s a win-win. We’re getting paid, why shouldn’t they? (PR4)

Finally, an additional incentive for participation could include reimbursement for transportation costs.

Data Collection—What, Where, and How

Patient and provider participants suggested a range of data to collect for a program of research on medical respite (see Table 1), where data collection should take place, and the need for data collection to be culturally safe, trauma-informed, and private.

Table 1.

Participant Reports on Data to Collect.

Data Category	Data Point	Data Source
Background information	Demographics (e.g., age, gender identity) Geographical region Shelter/housing status	Patient interview
Physical and mental health	Health event that led to hospitalization Physical functionality Health and follow-up care needs	Patient interview Hospital medical records
Health care and service utilization and costs	Hospital admissions Emergency department visits Length of hospital stay Connection to formal and informal supports Costs of services	Patient interview Hospital medical records
Outcomes of the medical respite intervention	Treatment effectiveness and appropriateness Health change Medication management Access to stable housing Quality of life Patient satisfaction with treatment	Patient interview

What data to collect

When asked what information would be important to collect as part of a program of research on medical respite, provider and patient participants suggested several key pieces: background information, including basic demographic information; the geographical region where patients come from or hope to go upon discharge from a medical respite; and their current shelter/housing status. In addition, when asked whether physical and mental health data should be collected from medical records, both participant groups agreed that this would be relevant information, so patients should be asked to give permission to access it. One patient (PA2) also suggested that researchers may need to obtain permission from health care organizations to access patient information: “I don’t know what the hospitals would say about that. You would need to get a release from the hospital, I guess.” The health and follow-up care needs of patients after discharge from a medical respite program were also identified as important data to collect.

Provider participants described the need to collect data on patients’ health care utilization, including hospital admissions, emergency department visits, and length of hospital stays, as well as patients’ connections to community services (e.g., outreach workers) or community teams. In addition, provider participants described the ability to use health care utilization data to estimate costs of a medical respite intervention program compared to costs of acute hospital stays.

Say they’ve been ready for discharge since October 1st, but they don’t get a [shelter] bed until the 10th—an additional 10 days that we’re paying over CA$1,000 a night for a [hospital] bed for these people. So, you could use that to calculate your cost savings, or your cost involved in it [compared to the cost of medical respite]. (PR4)

Another provider (PR2) reported that cost savings data are needed to effect policy change: “Ultimately, what funders care about is money and less service use. To be honest, I have not seen any changes in policy based on patient reported outcomes….”

Both participant groups suggested collecting outcome data, including patient’s quality of life, the effectiveness and appropriateness of the medical respite intervention (e.g., improved health or medication management or access to stable housing), and patient satisfaction with the intervention. A patient participant (PA15) offered some questions that future researchers could ask patients: “How did [the medical respite intervention] work out for you, essentially? Did we help you along your way to make your life easier to adjust into a more permanent lifestyle?”

Data collection locations

Participants indicated that researchers should go wherever potential participants are most comfortable for their first meeting. Once the researcher and participant have established rapport, they can work out new arrangements for follow-up interviews. Participants underscored that data collection locations need to be convenient and accessible for potential participants. They suggested that researchers go to the participants:

It’s better if the researcher goes to them and to create a safe environment within the space that they’re in because of transit costs or taxi costs or just even depending on where they are, how long it will take them to get to the appointment and then coming all the way back might take up quite a lot of their day. So those are the pieces [that can] become a great barrier for people. (PR10)

I would make it as easy as possible for them to be included. So, if that means going to them, in their room, with a consent form, and a pen—to make it as easy as pie, that’s what needs to happen. But if they have to jump through any hoops whatsoever, it just gets too overwhelming. (PA3)

Data collection should be culturally safe, trauma informed, and private

Regardless of what data are collected, participants highlighted that the approach to collection should be culturally safe, trauma informed, and kept private. While several patient participants mostly felt that nothing would be too personal to ask, they did indicate that other people may not agree. One patient participant (PA3) stated that no data would be too intrusive to collect from her perspective but also recognized that this may be different for people of other cultures:

Not for me. When I initially was hospitalized, maybe there would have been a bit of embarrassment. I still would have shared. There would have been some embarrassment about my situation…I do know that there were…[patients] of different cultures there and they did not discuss certain things. So, I think it’s [important to be] culturally sensitive.

Among the data too personal to collect, participants indicated that people might not want to be asked questions about stigmatized topics (e.g., history of sex offenses) or topics that would “trigger” painful memories of past trauma. A patient participant (PA4) stated, “It’s really knowing you’re not just talking to someone who has a physical injury, you’re talking to someone that has a physical injury that’s triggering and bringing up other stuff.” Such reports relate to the need for researchers to have empathy, be sensitive to participants’ histories, and engage with potential participants by building trust using a trauma-informed lens (Hopper et al., 2010). Patient participants suggested that “it probably comes down to how [the questions are] framed (PA4)” and “if the question is worded correctly, people would be more than willing to tell you because someone is listening, someone is expressing some sort of caring, they’re going to want to open up (PA10).” Other patient participants suggested that researchers clearly indicate to participants that they do not need to answer any questions that make them feel uncomfortable or else provide support to them during the interview. One patient participant (PA9) reflected on her previous participation in research and how these elements were built into the study:

Because there’s some pretty pointed questions that they ask in the [research] study, believe me, on women’s health and how certain medications affect us and a lot of the questions they ask, it’s quite personal…. But there is a disclaimer that says, “you do not have to answer these questions” or “if you’d like a counsellor present, we will provide one for you”….

Other participants indicated that if there is a question someone does not want to answer, they will likely let the researcher know and not answer it anyway.

Participants reported on the importance of having privacy checks in place in any proposed program of research on medial respite. Because some people might only be willing to share their personal information with certain individuals or organizations, potential participants need confidence in how personal data will be used and who will have access to it. The patient participant who had participated in another study (PA9) highlighted the confidence that participants had in sharing personal information because they were assigned anonymous and unique identifiers, which were stored in a separate location from the data:

Most of the women I know, they’re not afraid to answer any of it, [because they] were designated [anonymous] numbers, so that nobody out there can get that information because they don’t know who’s what number so you might want to designate…them a number and then add that name to a number in another ledger.

Although institutional research ethics boards require study investigators to maintain strict confidentiality and privacy of date, participants agreed that these protocols should be clearly outlined in any research proposal and communicated to potential participants during informed consent procedures.

Barriers and Motivation to Participate in Medical Respite Research

Participants identified both barriers to, and motivations for, participation in a program of research on medical respite. Potential barriers included (1) participants’ cognitive impairment, mental health, and use of substances; (2) participants not wanting to be involved with other people or not wanting to get in trouble or compromise themselves; and (3) language and cultural barriers. As one provider participant (PR2) stated, “One challenge with people who are older who are vulnerably housed and homeless is often cognitive impairment.” In addition, participants suggested that individuals might be hesitant to participate in research because they fear their involvement could jeopardize their welfare in some way, which might be a particular consideration for individuals who are vulnerably housed and reliant on others for housing and services.

Then again, you can get into trouble talking to some people—they would take what you said and use it against you somehow. I don’t even know that that’s realistic, but it’s possible…I wouldn’t like to be out on the street with one leg—you’re kind of vulnerable you know? (PA2)

Finally, willingness to participate may be affected by an individual’s cultural background or understanding of English and/or medical or research jargon; researchers should overcome these language and literacy barriers by changing the language used, as suggested by a patient participant (PA4):

Depending on their cultural situation, research language often can be beyond the scope of a lot of people in general…depending on their cultural background there might [also] be language barriers. So, are you able to do this research in languages other than English? And, also, just jargon in general—that needs to be tweaked somehow to meet the potential level of the individual. So, if they are from another cultural background and they are older, you have two variables there that might be challenging in terms of the jargon, [and] the academic language used in research often.

On the other hand, motivation to participate in a medical respite intervention study included (1) wanting to help oneself or others in a similar situation, (2) wanting to be heard and acknowledged, and (3) incentives. As summarized by a patient participant (PA2), people engage in research “for the betterment of everybody.” Another patient participant (PA3) agreed that they “want to help you help others like me.” Moreover, participating in research was identified as a way people can contribute and feel good about themselves. “It comes down to self-esteem. The more you help others, the better it is…(PA12)” Participants suggested that older adults may be especially apt to participate if they feel they can help improve their own circumstances or systemic problems.

Most of them actually want to help better themselves and anybody else that’s going to be in that situation. The older people actually see that the system is broken and they want to be a part of fixing it. They would love for people to hear them. (PR1)

The ability for research participants to have their stories heard or to engage in conversations with caring researchers was also highlighted as a potential motivation for research. As one provider participant (PR2) stated, “People want to be heard, right? Because a lot of times, no one is paying attention to what they want. Being listened to. Being asked questions. Getting their opinion.” Finally, additional motivations for participation in research suggested by both patient and provider participants included incentives such as money or gift cards or the belief that participation will lead to permanent housing.

Discussion and Applications to Practice

We explored three components of the feasibility of conducting an intervention study to test the effectiveness of a medical respite intervention for OPEH, a subpopulation of homeless patients not previously considered in the research literature. In the present study, older patients with lived experience of homelessness, as well as provider participants, endorsed the feasibility of an intervention study and provided valuable and meaningful suggestions to improve study design, access to patient participation, and data collection. Participants detailed what a program of research on medical respite should include in each of these areas, including qualities that researchers should have; preferences for study design; strategies for participant recruitment and retention; what, where, and how to collect data; and barriers and motivations to participation.

What emerged as especially salient in this study is the need for researchers to build trust with potential research participants in order to recruit and retain participants. By acknowledging that people who are experiencing homelessness are situated in vulnerable positions and may have long-standing mistrust of the health care system and persons in positions of power, researchers should demonstrate that participants are valued by consistently listening to and including them across all stages of research regardless of treatment arm. Previous research with other vulnerable groups, including mental health service consumers or youth who are experiencing homelessness, has similarly described the need to develop trust with research participants as a way to build mutual respect and understanding (Gaetz, 2014; Ochocka et al., 2002). Indeed, the process of trust building is especially important for longitudinal studies that hope to remain engaged with participants over time (Henwood et al., 2015).

“Meeting people where they are at” is a common sentiment that describes how practitioners should engage with people who are experiencing homelessness by providing them with services in the manner that they determine regardless of predetermined criteria that might preclude them (Noble, 2015). Participants suggested a similar philosophy be used by researchers when approaching potential research participants or collecting data. This included the need to be flexible in where data collection occurs and sensitive to how research questions are posed. As noted in prior peer-engaged research, being flexible in where data are collected contributes to feelings of safety (Greer et al., 2018). Implicit to this suggestion is the awareness that the investigative process may trigger past experiences of trauma, and therefore researchers must remain empathetic and sensitive to the mental and physical pain that patients may experience. Thus, as suggested for service delivery to people experiencing homelessness (Hopper et al., 2010), researchers should similarly utilize trauma-informed and culturally safe practices to avoid or mitigate such negative experiences (Canham et al., 2019; Goodwin & Tiderington, 2020). Trauma-informed research utilizes principles of cultural safety (including attending to intersections of participants’ cultural, historical, and gender experiences) and trauma-informed care (including safety, trustworthiness, choice, collaboration, and empowerment) to mitigate participant harm or upset and should apply these principles to all research—regardless of design—as a supplement to the ethical guidelines set forth by institutional review boards (Goodwin & Tiderington, 2020). Moreover, a trauma-informed research framework can be implemented throughout all stages of the research process (Voith et al., 2020).

A key strategy that participants suggested for building trust and engaging people who might be hesitant to share their experiences with researchers or others situated in traditional positions of power is to include peer researchers as part of the research team. This suggestion not only builds on the literature describing how to implement trauma-informed research methods but aligns with previous research that has benefited from the involvement of peer researchers in project development, data collection and interpretation, and knowledge dissemination (Greer et al., 2018; Ibáñez-Carrasco et al., 2019; Khandor & Mason, 2007). The unique contributions of peer researchers include the ability of peers to create a sense of security and to produce data that are more comprehensive as well as more relevant to communities that most benefit from the research knowledge (Greer et al., 2018; Ibáñez-Carrasco et al., 2019; Khandor & Mason, 2007). Such qualities align with trauma-informed research frameworks (Voith et al., 2020) and patient-oriented research methods (Barbic et al., 2019).

Discussions about the acceptability of patient randomization in a future RCT revealed several key points. While participants felt that it would be more equitable to provide everyone access to a medical respite intervention, they agreed that it would not be unethical to only serve some patients given limited available resources. Similar to previous research describing medical respite trials as no less unethical than the “current status quo” (Buchanan et al., 2006), participants indicated that a control group would reflect “business as usual” when the program was at capacity. Furthermore, in this scenario, the goal of the study would be to monitor the health and housing status of patients who were able to access medical respite (the experimental group), as well as those who were unable to access medical respite because it was at capacity (the control group). What also became evident is the perspective that participation in an intervention study on medical respite would benefit all participants regardless of their group location. This attitude indicates another motivational reason to be involved in this research and suggests that recruitment and longitudinal retention is achievable. Additionally, participants recommended follow-up interviews to add richness and depth to quantitative data, which can be another way for participants to feel they are making a meaningful contribution to change and thereby stay engaged in research over time.

In the development of a medical respite intervention study, participants emphasized the need for members of both the experimental and control groups to be followed longitudinally to compare long-term health and housing outcomes. While participants thought that it would be easier to stay engaged with patients who receive the medical respite intervention over time, participants also thought that even members of the control group would remain engaged if appropriately incentivized. Financial incentives were reported to be a motivator for ongoing engagement with researchers regardless of whether the patient was receiving an intervention. Although there have been questions about whether research incentives are coercive, Singer and Bossarte (2006) conclude that incentives are never coercive. Further, they suggest that by sampling members of the study population who are not part of the study sample about the appropriateness and amount of compensation, as done in the current study, researchers and institutional review boards can best evaluate risks and benefits (Singer & Bossarte, 2006). Thus, future research should use payment standards developed in collaboration with peers and providers to guide the selection and use of financial incentives (British Columbia Centre for Disease Control, 2018; Greer et al., 2018).

The ability to stay engaged with researchers was highlighted by most participants as potentially challenging for patients who have unstable housing or limited access to communication resources (i.e., phone, email). Suggestions for maintaining contact were offered, including getting contact information at the outset of a study about close others and identifying regular locations where participants spend time. Prior research with persons experiencing homelessness has cited similar challenges and suggested overcoming these tracking challenges by developing a comprehensive plan for maintaining regular contact—for instance, by asking participants to provide a list of people who might know how to find them as well as places where they frequent (E. H. Cohen et al., 1993; Goering et al., 2011). Prior longitudinal research has found that collecting comprehensive tracking data at baseline combined with an interviewer’s ability to develop rapport with study participants are key factors that enhance participant retention (Bootsmiller et al., 1998; E. H. Cohen et al., 1993; Gerlitz et al., 2017).

Data that participants felt would be of interest and relevance to collect in a medical respite study included patient demographic and background information; current and historical physical and mental health conditions; current health needs; health care and service utilization and associated costs; and quality of life, patient satisfaction, and health and housing outcomes following the medical respite intervention. As in previous research (Sadowski & Buchanan, 2009), participants of the current study suggested that data could be collected through in-person self-reports or verified through the review of medical records. While participants described a range of existing data systems that collect some of the data of interest, no single data source that could serve as a comprehensive database for analysis was identified. Instead, participants suggested that a new database be developed that drew on various sources of information, so that researchers could ensure relevant scope and quality of the data.

Ultimately, it was acknowledged that the research question asked in a medical respite study would inform the study design and methods, as well as participant eligibility criteria. However, regardless of the proposed research question, participants highlighted the need to ensure data confidentiality and patient privacy, especially given the vulnerability experienced by many OPEH. It was recognized that a research ethics board should provide oversight to a proposed study to help maintain confidentiality and privacy. Without these ethical considerations in place, participants recognized that potential participants would not be willing to participate or share personal information.

One limitation to this study is the potential that patient participants represent a biased perspective on the willingness of other people with lived experience to engage in research. That participants of this study volunteered to be interviewed suggests that they may be more open to sharing personal information than others. Indeed, most of the patient participants suggested that they would be willing to share any information about themselves while simultaneously acknowledging that others may be more hesitant. Second, following the first few patient participant interviews, we slightly modified the interview agenda because responses to several questions about how to design a future research study and what data to collect were deflected back to the interviewer because participants thought the researcher (as the professional) would have more knowledge about these topics. To better situate the researcher as a learner, the researcher then started each interview by acknowledging the inherent awkwardness that participants might feel in answering some of the questions and emphasized the value in researchers understanding the perspective of people who may have never designed or participated in a research study. Following this adjustment to how questions were prefaced, patient participants were more forthcoming with suggestions and ideas about the research design and process.

Despite these limitations, findings from this study provide important insight into how to design intervention research with marginalized populations, including whether medical respite has more individual (e.g., improved health outcomes and/or access to permanent housing) and system (e.g., cost effectiveness) benefits compared to treatment as usual (Zerger, 2006). Findings from this study build on an emerging evidence base (Goodwin & Tiderington, 2020; Voith et al., 2020) about trauma-informed research practices with marginalized groups by contributing the voices of OPEH to this dialogue. In addition, the inclusion of peer researchers on research teams is a noteworthy contribution to discussions on how to design trauma-informed research. For instance, peer researchers can be best positioned to increase recruitment and retention, as well as ensure that interview guides are understandable and that incentives are appropriate. As in the present study, patient partners should be included as advisors throughout the research process, including the development of the study proposal and questions on the interview guide.

Footnotes

Appendix A

Authors’ Note

The opinions and interpretations in this publication are those of the authors and do not necessarily reflect those of MSFHR or BC SUPPORT Unit.

Acknowledgments

We respectfully acknowledge that the research for this article was conducted on the unceded territories of the Coast Salish people, including the territories of the x^wməθkwəyəm (Musqueam), Skwxwú7mesh (Squamish), and Səlílwətaʔ/Selilwitulh (Tsleil-Waututh) Nations, on which we are privileged to live, work, and play.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was made possible by a Pathway to Patient-Oriented Research (P2P) award funded by the Michael Smith Foundation for Health Research (MSFHR) in partnership with the BC SUPPORT Unit.

ORCID iD

Sarah L. Canham

References

Barbic

S. P.

Leon

Manion

Irving

Zivanovic

Jenkins

Ben-David

Azar

Salmon

Helps

Gillingham

Beaulieu

Pattison

Talon

Oyedele

Tee

Mathias

(2019). Understanding the mental health and recovery needs of Canadian youth with mental health disorders: A strategy for patient-oriented research (SPOR) collaboration protocol. International Journal of Mental Health Systems, 13, 6. https://doi.org/10.1186/s13033-019-0264-0

Bauer

Moughamian

Viloria

Schneidermann

(2012). Leaving before discharge from a homeless medical respite program: Predisposing factors and impact on selected outcomes. Journal of Health Care for the Poor and Underserved, 23(3), 1092–1105. https://doi.org/10.1353/hpu.2012.0118

Bootsmiller

B. J.

Ribisl

K. M.

Mowbray

C. T.

Davidson

W. S.

Walton

M. A.

Herman

S. E.

(1998). Methods of ensuring high follow-up rates: Lessons from a longitudinal study of dual diagnosed participants. Substance Use & Misuse, 33(13), 2665–2685. https://doi.org/10.3109/10826089809059344

British Columbia Centre for Disease Control. (2018, February). Peer payment standards. 1–15. http://www.bccdc.ca/resource-gallery/Documents/EducationalMaterials/Epid/Other/peer_payment-guide_2018_printer-friendly.pdf

Boyatzis

R. E.

(1998). Transforming qualitative information: Thematic analysis and code development. ISBN: 9780761909613

Braun

Clarke

(2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3, 77–101. https://doi.org/10.1191/1478088706qp063oa

Brown

R. T.

Hemati

Riley

E. D.

Lee

C. T.

Ponath

Tieu

Guzman

Kushel

M. B.

(2017). Geriatric conditions in a population-based sample of older homeless adults. The Gerontologist, 57(4), 757–766. https://doi.org/10.1093/geront/gnw011

Buchanan

Doblin

Sai

Garcia

(2006). The effects of respite care for homeless patients: A cohort study. American Journal of Public Health, 96(7), 1278–1281. https://doi.org/10.2105/AJPH.2005.067850

Buchanan

Kee

Sadowski

L. S.

Garcia

(2009). The health impact of supportive housing for HIV-positive homeless patients: A randomized controlled trial. American Journal of Public Health, 99(suppl 3), 675–680. https://doi.org/10.2105/AJPH.2008.137810

10.

Canham

S. L.

(2019). Determining the feasibility of a medical respite intervention study for older homeless patients in Metro Vancouver. https://bcsupportunit.ca/system/files/resources/2019-05/P2Preport_Determiningthefeasibilityofmedicalrespite_Final_0.pdf

11.

Canham

S. L.

Custodio

Mauboules

Good

Bosma

(2020). Health and psychosocial needs of older adults who are experiencing homelessness following hospital discharge. The Gerontologist, 60(4), 715–724. https://doi.org/10.1093/geront/gnz078

12.

Canham

S. L.

Davidson

Custodio

Mauboules

Good

Wister

A. V.

Bosma

(2019). Health supports needed for homeless persons transitioning from hospitals. Health and Social Care in the Community, 27(3), 531–545. https://doi.org/10.1111/hsc.12599

13.

Canham

S. L.

Humphries

Danielsen

Small

Bosma

. (in press). Design considerations for the development and implementation of a medical respite for older adults experiencing homelessness in Metro Vancouver. Medical Care.

14.

Ciambrone

Edgington

(2009). Medical respite services for homeless people: Practical planning. https://nhchc.org/wp-content/uploads/2019/08/FINALRespiteMonograph1.pdf

15.

Cohen

C. I.

(1999). Aging and homelessness. The Gerontologist, 39(1), 5–14. https://doi.org/10.1093/geront/39.1.5

16.

Cohen

E. H.

Mowbray

C. T.

Bybee

Yeich

Ribisl

Freddolino

P. P.

(1993). Tracking and follow-up methods for research on homelessness. Evaluation Review, 17(3), 331–352. https://doi.org/10.1177/0193841x9301700305

17.

Cornes

Whiteford

Manthorpe

Neale

Byng

Hewett

Clark

Kilmister

Fuller

Aldridge

Tinelli

(2018). Improving hospital discharge arrangements for people who are homeless: A realist synthesis of the intermediate care literature. Health & Social Care in the Community, 26, e345–e359. https://doi.org/10.1111/hsc.12474

18.

Doran

K. M.

Ragins

K. T.

Gross

C. P.

Zerger

(2013). Medical respite programs for homeless patients: A systematic review. Journal of Health Care for the Poor and Underserved, 24(2), 499–524. https://doi.org/10.1353/hpu.2013.0053

19.

Dorney-Smith

Hewett

Burridge

(2016). Homeless medical respite in the UK: A needs assessment for South London. British Journal of Healthcare Management, 22(8), 405–413. https://doi.org/10.12968/bjhc.2016.22.8.405

20.

Eldridge

S. M.

Lancaster

G. A.

Campbell

M. J.

Thabane

Hopewell

Coleman

C. L.

Bond

C. M.

(2016). Defining feasibility and pilot studies in preparation for randomised controlled trials: Development of a conceptual framework. PLoS One, 11(3), 1–22. https://doi.org/10.1371/journal.pone.0150205

21.

Gaetz

. (2014). A safe & decent place to live: Towards a housing first framework for youth. Canadian Observatory on Homelessness Press website: http://homelesshub.ca/sites/default/files/HFFWYouth-FullReport_0.pdf

22.

Gerlitz

Joyce

Gadermann

A. M.

Hubley

A. M.

Ma Zhu

Stowe

Palepu

(2017). Observations from the field: Interviewers’ experiences interviewing and retaining homeless and vulnerably housed participants in longitudinal health research. Journal of Social Distress and the Homeless, 26(2), 128–137. https://doi.org/10.1080/10530789.2017.1349598

23.

Goering

P. N.

Streiner

D. L.

Adair

Aubry

Barker

Distasio

Hwang

S. W.

Komaroff

Latimer

Somers

Zabkiewicz

D. M.

(2011). The at home/Chez Soi trial protocol: A pragmatic, multi-site, randomised controlled trial of a housing first intervention for homeless individuals with mental illness in five Canadian cities. BMJ Open, 1(2), 1–18. https://doi.org/10.1136/bmjopen-2011-000323

24.

Goodwin

Tiderington

(2020). Building trauma-informed research competencies in social work education. Social Work Education. https://doi.org/10.1080/02615479.2020.1820977

25.

Greer

A. M.

Amlani

Pauly

Burmeister

Buxton

J. A.

(2018). Participant, peer and PEEP: Considerations and strategies for involving people who have used illicit substances as assistants and advisors in research. BMC Public Health, 18(1), 1–11. https://doi.org/10.1186/s12889-018-5765-2

26.

Henwood

B. F.

Stefancic

Petering

Schreiber

Abrams

Padgett

D. K.

(2015). Social relationships of dually diagnosed homeless adults following enrollment in housing first or traditional treatment services. Journal of the Society for Social Work and Research, 6(3), 385–406. https://doi.org/10.1086/682583

27.

Hopper

E. K. E.

Bassuk

E. L.

Olivet

(2010). Shelter from the storm: Trauma-informed care in homelessness services settings. The Open Health Services and Policy Journal, 3(617), 80–100. https://doi.org/10.2174/1874924001003020080

28.

Ibáñez-Carrasco

Watson

J. R.

Tavares

(2019). Supporting peer researchers: Recommendations from our lived experience/expertise in community-based research in Canada. Harm Reduction Journal, 16(1), 1–5. https://doi.org/10.1186/s12954-019-0322-6

29.

Khandor

Mason

. (2007). The street health report 2007. https://www.streethealth.ca/downloads/the-street-health-report-2007.pdf

30.

National Health Care for the Homeless Council. (2012). Medical respite care and homelessness: 2012 policy statement. https://nhchc.org/wp-content/uploads/2019/08/2012-Medical-Respite-Policy-Statement-doc.pdf

31.

National Health Care for the Homeless Council. (2013). 2013 Medical respite program directory: Descriptions of medical respite programs in the United States and Canada. http://docplayer.net/33871235-2013-medical-respite-program-directory-descriptions-of-medical-respite-programs-in-the-united-states-and-canada.html

32.

Noble

(2015). Beyond housing first: A holistic response to family homelessness in Canada. https://homelesshub.ca/sites/default/files/attachments/2015_HousingFirstReport_EN-WEB.pdf

33.

Ochocka

Janzen

Nelson

(2002). Sharing power and knowledge: Professional and mental health consumer/survivor researchers working together in a participatory action research project. Psychiatric Rehabilitation Journal, 25(4), 379–387. https://doi.org/10.1037/h0094999

34.

QSR International. (2018). NVivo qualitative data analysis software (Version 12). NVivo. http://www.qsrinternational.com

35.

Sadowski

L. S.

Buchanan

(2009). Effect of a housing and case management program on emergency department visits and hospitalizations among chronically ill homeless adults. JAMA, 301(17), 1771–1778. https://doi.org/10.1001/jama.2009.561

36.

Singer

Bossarte

R. M.

(2006). Incentives for survey participation. When are they “coercive”? American Journal of Preventive Medicine, 31(5), 411–418. https://doi.org/10.1016/j.amepre.2006.07.013

37.

van Laere

de Wit

Klazinga

(2009). Shelter-based convalescence for homeless adults in Amsterdam: A descriptive study. BMC Health Services Research, 9(1), 208. https://doi.org/10.1186/1472-6963-9-208

38.

Voith

L. A.

Hamler

Francis

M. W.

Lee

Korsch-Williams

(2020). Using a trauma-informed, socially just research framework with marginalized populations: Practices and barriers to implementation. Social Work Research, 44(3), 169–181. https://doi.org/10.1093/swr/svaa013

39.

Zerger

(2006). An evaluation of the respite pilot initiative. http://www.nhchc.org/wp-content/uploads/2011/09/RespiteRpt0306.pdf

40.

Zerger

Doblin

Thompson

(2009). Medical respite care for homeless people: A growing national phenomenon. Journal of Health Care for the Poor and Underserved, 20(1), 36–41. https://doi.org/10.1353/hpu.0.0098

41.

Zlotnick

Zerger

Wolfe

P. B.

(2013). Health care for the homeless: What we have learned in the past 30 years and what’s next. American Journal of Public Health, 103(S2), S199–S205. https://doi.org/10.2105/AJPH.2013.301586