2.1. Methodology

This analysis is part of a larger ethnographic study titled, “Caregiving for vulnerable and marginalized older adults at the end of life” (CIHR, PJT—173369), which is taking place in three Canadian cities (Oct 2021–Sept 2025) and aims to characterize caregiving at the end of life for people experiencing structural vulnerability⁵. Our research teams’ interests in family and caregiving stemmed from over a decade of research exploring end-of-life experiences of people facing health and social inequities [32, 35, 42]. The current analysis, led by the lead author (Ashley Mollison), represents one analysis out of this larger study and comprises part of Ashley Mollison’s PhD dissertation.

Our methodological approach is ethnography informed by critical theoretical perspectives and integrates social justice and health equity approaches. While there is no one single definition, all ethnographies tend to have a common research approach, which centers on people in their “natural” context, employing unstructured data collection and a range of data sources, and utilizes smaller sample sizes and in-depth cases in order to understand how people make sense of their world [43–45]. Unlike some ethnography that tends to present deep, apolitical descriptions, critical ethnographies recognize the political nature of health and social institutions and the ideologies that underlie them [45]. Critical approaches lend themselves well to social justice aims to both understand and address how power embedded in institutional and social processes makes people vulnerable to health risk and poor health outcomes [42, 46, 47]. Critical ethnography involves turning the gaze toward research itself and being reflexive about how power and discourse shape what we see in the field and what we interpret as truth [44]. These approaches encourage not just critical examination of structures and systems but challenge researchers to unpack their assumptions and be critically reflexive in the research process to avoid further harming participants already facing structural violence [48].

Social justice and health equity approaches encourage researchers to understand health and social issues as a product of social and structural inequalities [42, 49]. Rather than seeing differences in health access and outcomes between groups of people as “natural” or the result of genetics or behavior, these differences are recognized as unfair and unjust because they rely on systemic devaluating and privileging of groups of people [50]. Racism, sexism, homophobia, ableism, and other processes embedded in economic, political, health, and social structures mean that people do not have the same access to the conditions for good health [42, 50]. Despite (or in spite of) these inequities, people and communities facing historical and ongoing oppression have developed strategies for survival that we have a lot to learn from [51]. In relation to our area of study, caregiving is understood within a broader context of undervalued “care” work that is gendered, racialized, and classed [52]. Yet, people continue to care for each other and push back against dominant ideology, concepts, and approaches that define what care is and who the caregivers are or should be.

2.2. Data Collection

Research partners include equity-oriented palliative teams in three cities whose clinical practice focuses on people facing poverty, homelessness, racism, criminalization, and discrimination along with life-limiting conditions. Prior to data collection, several meetings were held between research team members and community partners to codevelop research processes (e.g., study advertisement and participant recruitment). The research team shaping and carrying out the observations all identified as White women, housed, and with financial and social resources to render us systemically privileged. The lead author has held various roles over the last 15 years in relation to homelessness such as a volunteer, community organizer, caregiver, worker, and researcher. In planning stages and during the data collection period, several scheduled meetings and unscheduled debriefs were held among members of the data collection team in an effort to critically analyze issues of power and privilege and how those experiences shaped data collection, analysis, and thematic development. Once the study received ethics approval (H22-00313 (BC/ON); HREBA. CHC-22-0032 (AB)), data collection was initiated with three main groups: members of equity-oriented palliative teams (e.g., physicians, nurses, psychiatrists, and outreach workers), clients and their informal caregivers, and other formal (paid) caregivers (e.g., shelter/housing workers, support/outreach workers, and case managers)⁶.

2.3. Equity-Oriented Palliative Teams

In-depth interviews with members of the equity-oriented palliative teams were completed to seek insights on their everyday work in relation to family and caregiving (see Interview Guide in Supporting Information (here)). These interviews lasted approximately 60 min and were recorded and transcribed verbatim. Following interviews and preliminary analysis, we began attending clinical rounds where teams discussed client cases including care coordination, system barriers, relational challenges, and broader issues around care and support. After observing rounds for several months, we observed team members in their work settings (e.g., shelters, housing units, health units, and on the streets). Fieldnotes were handwritten and then transcribed. We focused on the settings, activities, acts (what people were doing), events, and time and included reflexive notes that captured broader concepts around family and caregiving, how our positionality as researchers shaped observations [48], as well as lines of enquiry for future data collection.

2.4. Clients and Caregivers

The equity-oriented palliative teams were asked to identify and, with their consent, introduce us to clients who may have a “somebody” in their life who was offering care or support to them. It was emphasized that these people did not need to be biolegally related nor considered by the client or supporter to be a “caregiver.”⁷ Once the connection with the client was made, we obtained written, informed consent and prioritized relationship building. We met with participants in various settings including health clinics and social service agencies, outdoors, restaurants/coffee shops, and their homes. Engagement also included texts and phone calls that were documented in in-depth fieldnotes. As relationships were built with the client, we sought permission to engage with others in their life who were providing them with some sort of support. Depending on the comfort and permissions from the client, we met with the caregiver either together with, or separately from, the client. Occasionally, we were connected with a caregiver after the death of a client so, in these cases, we engaged solely with the caregiver. We engaged in longitudinal data collection (observations and interviews) with clients and caregivers approximately once every 4–6 weeks for 15 months (July 2023–November 2024), but engagement fluctuated including intensifying in periods of crisis or if things were changing quickly for people (e.g., functional decline). Observations lasted no longer than 2 h at a time. We often recorded conversations and then transcribed them into fieldnotes resulting in verbatim quotes. All clients and caregivers were offered a $50 cash honorarium per visit (observation or interview). Honorariums were often supplemented with coffee and food to show appreciation for the time that people took to participate in research and were also aligned with the justice-oriented approach in being responsive to food insecurity for many clients/caregivers.

2.5. Other Service Providers

Other service providers were found to be involved in caregiving and support during our observations with clients and caregivers. In these cases, we sought consent for observations. In other cases, clients and caregivers identified service providers in their circle of care and support and we thus obtained consent and conducted in-depth interviews with these providers (see Interview Guide in Supporting Information (here)). Interviews followed the same interview guide as equity-oriented palliative teams, were approximately 60 min, and were recorded and transcribed verbatim. Service providers were offered a $50 cash honorarium per interview if they were participating in research outside of their work time.

While the larger study is ongoing, this analysis is based on data collected over a 27-month period (August 2022–November 2024). During this study period, approximately 300 h of observation and 44 interviews were conducted with members of equity-oriented palliative care teams (n = 20), clients living with life-limiting conditions (n = 8), caregivers (n = 10), and service providers (n = 44).

2.6. Analysis

Consistent with an ethnographic approach, we utilized reflexive thematic analysis [53] to generate themes. The analysis began by purposefully selecting two client cases that illustrated diverse caregiving scenarios. The in-depth cases included instances where the individual’s situation was discussed in clinical rounds, observations and interviews with the person, their caregiver, and service providers. An analytic subgroup composed of three data collectors (Ashley Mollison, Alexandra Stewart, and Amber Bourgeois) and two members of the broader research team (Marilou Gagnon and Kelli I. Stajduhar) read the compiled data and independently developed a “big bucket” coding scheme. The core analytic team discussed and came to a consensus on the development of the preliminary coding scheme. This coding scheme was tested, revised, and refined several times until producing a final scheme that was then thematically applied to the entire dataset of interviews and observational fieldnotes using NVivo 14 qualitative data analysis software. Overtime, through observations and interviews, patterns began to emerge that were tested and reformulated utilizing inductive and deductive analysis [54]. Data on remaining clients and caregivers were compiled into cases and read repeatedly to inform the thematic analysis. Coding of the remaining data was conducted (Ashley Mollison, Alexandra Stewart, and Amber Bourgeois), supported by the broader analytic team. The analysis in this paper stems from a two-bucket code called “Family” and “Seeing and engaging the caregivers.” Themes were developed and interrogated by the broader research team and community partners, then revised, and refined.

3.1. Relationships and Roles in Informal Caregiving

Of the 10 caregivers included in this study, 6 were friends, 3 were spouses/partners or ex-partners, and 1 was a biological sibling. Despite this categorization, there was fluidity in the nonbiolegal relationships. In some cases, people drew on concepts of love, support, and even family-like terms to define their unrelated relationships. Many friend relationships were longstanding and shifted with the requirement for care. For instance, Sasha was a childhood friend of Russ, a 61-year-old White man, and became the primary caregiver when he was diagnosed with advanced lung cancer while homeless. Sasha provided instrumental and emotional support to Russ including transportation and accompaniment to medical appointments, housing him and his dog for periods of time, food preparation, social outings, light housekeeping, and agreeing to be his decision-maker; yet, their relationship was complicated and “on and off” during the year and a half our team spent engaged in Russ’ life. They eventually had a falling out and ended their friendship as Russ declined and approached the end of life.

Mark, a 59-year-old White man with advanced COPD and schizophrenia, and Robert, a White man in his 50s, lived in separate units in the same subsidized housing facility. Mark shared that while he had known Robert for 20 years, “the caregiving part of their relationship was quite recent due to his declining health” (Fieldnote). While Mark had access to formal palliative care services, Robert stepped up to provide more support with personal, social, and instrumental care as Mark declined including companionship, food and supply runs, and cooking.

Dwayne associated care with “just being human” and saw it as making up for his troubled past. On the day of Bill’s planned MAiD death, they sat together all day, telling stories, and listening to music until Bill died.

Rather than distinct caregiver or care recipient roles, informal caregiving in this context reflected more of a community care model, where various people (including the person living with a life-limiting condition) engaged in reciprocal care exchanges, helping each other out whenever a need arises. These participants often drew on their past or lifetime experiences of care to describe relationships, approaches, and motivations for care. Carla, a 53-year-old Indigenous woman, described her care for a man in his 70s, a former shelter worker whom she met when she was homeless. She went to his place 1-2 times per week, emptied the garbage, did his dishes, swept his floor, and did his laundry. Rather than unidirectional care, Carla described a “give and take” in a relationship that had initiated as service provider/client but developed into a friendship over the years.

The case of Ronald, a White man, and David, an Indigenous man, both in their 60s, illustrated another mutual care relationship without a clear distinction between caregiver and care recipient. Ronald lived with multiple comorbidities and was labeled “failure to thrive” by his palliative care team, while David lived with mental health and substance use issues resulting in ongoing falls and hospitalizations. An outreach worker told us that they were “best friends” and that they met each other in their subsidized housing facility. Beginning as “drinking buddies,” their relationship grew as they visited each other in the hospital, provided each other emotional support, and shared the costs of groceries and supplies. During an observation, David told us, “I look out for him, and he looks out for me.”

Far from being the sole caregiver in the palliative context, we witnessed informal caregivers supplementing care by formal workers (e.g., palliative teams, housing workers, and case managers) to provide check-ins, supply runs, conversation, and emotional and mental health support to one another.

3.2. Identifying and Perceiving Informal Caregivers

Service providers supporting people in contexts of inequity acknowledged that there is a lack of understanding and visibility of informal networks of support in mainstream healthcare. According to a palliative care physician, “As someone who works in mainstream healthcare, I will tell you that often there is an absolute lack of recognition of the street family caregiving network.” Service providers went further to report instances of discrimination and stigma that friends and street family experience when attempting to carry out caregiving roles such as accompaniment or visits in mainstream health settings.

Beyond mainstream health settings, barriers to caregiving were identified in community settings. Two caregivers faced barriers accessing the housing units of the people they were caring for due to policies preventing them from holding another resident’s keys. While Sasha lamented the fact that building management would not provide her with a key to her friend’s suite, Dwayne took it upon himself to hold his friend’s keys against the wishes of management creating tension between him and his housing provider. Another caregiver was banned from the housing and healthcare facility where her spouse was living, forcing them to see one another outside the building. These examples illustrate how policies and practices can contribute to a lack of recognition and significance of caregiving roles.

This incident speaks to the ways in which caregiving can be kept intentionally invisible for fear of exposing vulnerability or of eviction from community settings due to higher medical needs.

While many service providers recognized the value of “street family,” a common theme was the concern about the stability of these relationships, the nature of the care provided, and the impact of this caring work on the caregiver and the person living with life-limiting conditions.

Beyond service providers expressing concern at these kinds of situations, our analysis also revealed instances of institutions mobilizing policy and practices to respond and separate people from their chosen caregivers. We also saw examples of institutions making exceptions to policy and practice (e.g., guest policies) to enable connection to caregivers if the person was perceived as helpful or beneficial to the client.

3.3. Biolegal Privileging and Exclusion of Unrelated Caregivers

In contrast to dominant characteristics ascribed to family (e.g., care, love, loyalty, and support), our participants shared stories of harm and violence from their families (e.g., families of origin, foster care families, and adopted families). Russ, for instance, shared with us his poor lifetime experiences with family including abuse in childhood, attempted and failed family connections, and the implications of this on his end-of-life wishes. None of his biolegal family members lived close to Russ, nor visited, nor provided personal, instrumental, or emotional care to him at the end of life. Russ had elected to do MAiD and indicated that he did not want his family to attend. Instead, he received support from formal palliative care services, members of an equity-oriented palliative care team, a volunteer, friends such as Sasha, and neighbors.

While biolegal families were rarely present for everyday care, people living with life-limiting conditions were not often fully estranged from their family members. In a palliative context, some participants received calls from family; others expressed knowing that they could call family if they were in need. In some cases, the desire to connect with family grew as the person approached the end of life highlighting how, despite estrangement, family reconnection and reconciliation can, for some people, become more important overtime.

When we met Donna, a 58-year-old White woman with advanced lung cancer, and her partner, Max, a 64-year-old White man, they were living at a shelter while she accessed daily cancer treatment. Out of a large family that lived in the province, Donna was only in communication with one of her three children but expressed being able to call her family if she really needed them for “moral support.” As Donna got sicker, reconnecting with family became more important to her, and her equity-oriented palliative team supported Donna and Max to attend her family reunion before she died.

While these informal caregivers sometimes felt that “stepping back” was appropriate, our findings suggest that more commonly, previously estranged or uninvolved biolegal family not only came out of the woodwork at the end of life but also “took over” care. When this occurred, unrelated caregivers felt excluded, even though they were important to people living with life-limiting conditions and their care. One case presented conflict between an estranged family and Jacob, a 7-year-old man and close friend of Roxana, a 55-year-old woman who died of advanced cancer. While Roxana wanted Jacob to be her executor, he was not legally documented as such. When Roxana was admitted to the hospital at the end of her life, the hospital staff called her estranged mom and sister following the next of kin policy and legislation.

As this case illustrates, reunification was not always a choice of the person or their family but rather a consequence of family members being considered next of kin or listed as emergency contact activated by health, palliative, and legal systems. Roxana died in the hospital, and the hospital staff had contacted her biolegal family to begin postdeath arrangements.

We found that unrelated people took on important care roles but faced challenges being legitimized as informal caregivers. To mitigate some of these challenges, equity-oriented palliative care teams did what they could to support friends and chosen family who stepped in to provide care at the end of life. In the case of Jacob, the equity-oriented care team supported him emotionally, with a housing application, and held some of Roxanna’s belongings while Jacob was unhoused. For Dwayne, the team checked in with him during the end of life and death of his friend, Bill, and supported him with postdeath arrangements. Where people had spouses/partners involved in their care, equity-oriented palliative teams advocated for housing and income support for both people; in the case of Ronald and David, the team advocated for them to be housed together in the same assisted living facility. These teams faced challenges, however, in their support for caregivers due to constraints on time and resources and a lack of access to legal and financial support to legitimize the role of unrelated caregivers for participation in end-of-life and postdeath care.

4.1. Implications for Policy, Practice, and Research

There are increasing commitments to meet the palliative needs of underserved communities [4, 37, 39] alongside a recognition of how older adults are expending their networks beyond biolegal family for care and support [14–17, 58]. Additional attention and innovative approaches will be required to see, engage, and support unrelated, informal caregivers, many of whom face similar Social Determinants of Health needs as the people they care for alongside discrimination, racism, health conditions, and disabilities [26]. Compassionate community approaches [59, 60] have been suggested as a key strategy for meeting the palliative needs of more people [37, 41] and inspiring examples exist of projects that draw on the inherent strengths of community to care for each other in contexts of inequity and serious illness [61]. Despite challenging and estranged relationships in life, the end-of-life phase can present the conditions for family reunification [26, 62]. Our findings suggest a need to build capacity for trauma- and violence-informed care [63] and culturally safer and antiracist approaches [64–67] in palliative care to adequately attend to relationships and histories with family and community disconnection and reconnection (see also [68, 69]). Supporting the “caregiver constellation” will require supportive policies and practices that consider and mitigate diverse caregiver needs given differential access to power and resources. Future research may investigate where this phenomenon has happened before (e.g., among the chosen families caring for those dying during the AIDS epidemic) and how exclusions of relationships that are not biologically or legally sanctioned leave lasting adverse impacts on caregivers such as disenfranchised grief [70, 71]. Largely unexplored in the study was the experiences of biolegal family, previously estranged, who came on the scene at the end of life or postdeath, which is an area of future study.