Understanding how social context shapes decisions to seek institutional care: a qualitative study of experiences of progressive cognitive decline among Latinx families

Abstract

Background and Objectives

Latinx individuals in the United States experience elevated risk of cognitive impairment, less access to institutional care, and more caregiver strain. Cultural orientations have been emphasized as a factor in caregiving decisions in Latinx families but rarely in the context of community structural position and institutional circumstances. This ethnographic case study illustrates how cultural, institutional, and structural factors shape Latinx families’ deliberations and decisions about institutional care for members experiencing cognitive decline.

Research Design and Methods

We use comparative ethnography to examine experiences of older adults with cognitive decline and their care partners. Analysis focuses on ten care partners and nine older adults in urban and rural Northern California. Drawing on analytical memos that integrate data from semi-structured interviews and participant observation in community settings, we explore how older adults and care partners make decisions to pursue institutional care.

Results

Among our Latinx respondents, decisions to relocate from private to institutional care is complex and context-dependent. These decisions reflect cultural orientations including values of familism and individuality as well as access to resources that reflect structural circumstances and support networks that assist in navigating local institutions. Familism is enacted in diverse ways, with family involvement playing a role before, during, and beyond institutionalization.

Discussion and Implications

This study illustrates how social factors contextualize culturally mediated decisions regarding older adult caregiving in Latinx families. Its results imply that navigational assistance could support decision-making in Latinx communities. It highlights the utility of fine-grained data for accurately capturing this dynamic process.

Background and objectives

People living with progressive cognitive decline such as Alzheimer’s Disease and related dementias (ADRD) and their care partners face numerous decisions during the course of illness, including decisions about diagnosis and treatment (Bradford et al., 2009; Xanthopoulou & McCabe, 2019), decisions about lifestyle and safety at home (Carr et al., 2005; Perkinson et al., 2005; Tudor Car et al., 2017), and decisions related to end of life (Harrison et al., 2022; Konno et al., 2024; Lassell et al., 2022). Highly consequential and complex decisions occur as illness progresses and older adults living with ADRD and their care partners face choices about living arrangements: bringing in-home care support (Ritchie et al., 2018), arranging co-residence with family members (Prickett & Angel, 2017), or relocating to institutional care (Teng et al., 2020).

This study examines the temporal, structural, and cultural factors that inform decision-making about institutional care among families of older Latinx individuals. We draw on ­Quesada’s pioneering conceptual work on structural vulnerability as applied to health and medicine. This provides leverage to understand the specific conditions and constraints within which Latinxs in the United States (US) grow old and allows us to appreciate how cultural orientations toward family obligation (Herrera et al., 2008) operate within contemporary social conditions of structural vulnerability in the US (Quesada et al., 2011). Our research goal is to understand how Latinx families value different kinds of living arrangements, what actions and responses are seen as appropriate within Latinx families given these values, and the living arrangement outcomes that prove possible in everyday life.

Their medical service needs and utilization mean US-residing Latinx face complex decisions about institutional care. Although they tend to live longer than other ethnic groups (Prickett & Angel, 2017), Latinx individuals bear the burden of higher levels of later-life dependency due to diabetes, hypertension, and functional limitations (Herrera et al., 2012), and lower levels of socio-economic standing, wealth, and health insurance coverage (Oney et al., 2022). Latinx have higher age-adjusted prevalence of dementia—with different estimates across studies and subgroups (Alzheimer’s Association, 2024; Graves et al., 2024; Hudomiet et al., 2022; Márquez et al., 2024)—yet lower service utilization rates (Hinton et al., 2024; Ornstein et al., 2022). Longitudinal analysis of the Hispanic Established Population for the Epidemiological Study of the Elderly (H-EPESE) data by Prickett and Angel (2017) found that declining physical and cognitive functioning and worsened financial strain are significant predictors of transitions in living arrangements among older Mexican Americans—even after accounting for individuals’ baseline levels of impairment. Their findings underscore the importance of time-varying health and economic factors in long-term care planning.

The social position of Latinx populations—quantitative analysts often use nation of origin (nativity) and length of stay in the US (acculturation) as measures of social position—also shapes decisions about institutional care. Analysis of H-EPESE by Herrera and colleagues (2012) examined living arrangement preferences among unmarried older Mexican Americans in wave 6 of the study. They investigated delays in institutionalization and argued that familism—defined as the socialization of grown children into traditional cultural norms of caregiving for aging relatives—continued to shape residential preferences, particularly among the Mexican-born. However, migration, financial strain, and workforce participation could limit families’ ability to follow culturally-preferred caregiving practices (Herrera et al 2012; see also Angel et al., 2004). Foreign-born Mexican American older adults have tended to remain more reliant on family support and government assistance; this reflects both structural constraints and cultural values (Angel et al., 1996). Among more acculturated, US-born Mexican Americans, there has been a gradual shift toward greater use of institutional care services. This trend reflects an increased ability of some families to afford and navigate the institutional care system (Herrera et al., 2012) at the same time as decreasing fertility and family migration limit the caregiving capacity of family-based Mexican American support networks (Angel et al., 2000). Herrera and colleagues (2008) found that social support, quality of healthcare referrals, and service knowledge influenced institutional care use among Mexican Americans living with ADRD. Acculturation was associated with higher education and income among Mexican families, but this had mixed effects on the provision of institutional care as higher income limited Medicaid eligibility but did not necessarily make private care affordable (Herrera et al., 2008). Gaugler and colleagues (2006) drew on Medicare Alzheimer’s Disease Demonstration Evaluation (MADDE) data to reveal that Latinxs tend to institutionalize older adults with dementia later than other ethnic groups in the US, but Medicaid eligibility increased the likelihood of institutionalization. Latinxs in this study mostly lived in Florida with a migration history from the Caribbean rather than Mexico (as in H-EPESE). These studies collectively reinforce the importance of considering financial access to care in addition to functional need when explaining care transitions among Latinx. These findings also highlight the diversity of the US-based Latinx population and the need to attend to the specific social circumstances of migration when examining health decision making.

Studies of the relationship between culture and Latinx health decisions have produced paradoxical results. The sociocultural stress model argues that ethnic identity can help account for caregiving outcomes among Latinxs (Aranda & Knight, 1997). Latinx families may lean on group-specific cultural understandings to reframe the meaning of care tasks (appraisal of burden) or draw on religious or co-ethnic support networks to provide distraction and alleviate the psychological burden of caregiving (Heo, 2014). A study examined the relationship between ethnicity and time to institutionalization among Latina (most of them Mexican-American or from the Caribbean Islands) and non-Latina White female dementia-caregiving dyads who participated in an intervention to reduce caregiver stress (Mausbach et al., 2004). Latinas in that study were less likely to institutionalize their relatives over an 18-month period, and positive views of caregiving helped explain Latinas’ decision to delay seeking institutional care. In other studies, however, the relationship between culture and caregiving decisions is less direct. Losada and colleagues (2010) show that individuals who endorse familism but face circumstances that make it difficult to provide care may experience higher levels of caregiver distress. This suggests that cultural orientations may create unrealistic expectations and avoidant coping mechanisms that exacerbate caregiving burdens and caregiver suffering (Losada et al., 2010). Kim et al note that familism “may represent obligation more than positive feelings about family support” (Kim et al., 2007, p. 573). Qualitative research has shown that caregivers may experience strain because traditional gender socialization may limit family support (Herrera et al., 2008) or because financial issues make family connections a source of conflict not support (Turner et al., 2015). Latinx caregivers of older adults living with dementia in Florida who experienced higher emotional burden (depression), provided more hours of care, or relied more on adult day services tended to more quickly pursue institutional care (Gaugler et al., 2006). The realities of caregiving, particularly the challenges facing sole caretakers who cannot rely on other family members, mean that the real-world impact of cultural familism may be difficult to ascertain.

Many scholars recognize the complexity of Latinx caregiving decisions, which are informed by disease status, family support capacity, structural conditions, and cultural orientations (Herrera et al., 2012). To advance our understanding of how Latinx families make everyday decisions about living arrangements, sociological theories of culture offer a useful tool. Sociologists have developed a rich literature—including practice theory and accounts of culture in action—to examine how culture shapes everyday decisions and actions undertaken in circumstances of physical, institutional, and social structural constraint and opportunity; this literature extends to decisions and actions surrounding aging and cognitive decline (Abramson, 2015; Bourdieu, 1977; Freidenberg, 2000; Swidler, 2023). These theories are adept at guiding the interpretation of qualitative data in general and the interpretation of comparative ethnographic data in particular. Ethnographic analysis can reveal how care is conceptualized and enacted and how the meaning of care interacts with, and is shaped by, structural circumstances and bureaucratic constraints. Comparative data within groups that face shared experiences—such as structural vulnerability and racialization in the case of Latinx residents in the US—is particularly valuable for understanding when families avoid institutionalization as well as the decisions of those who pursue institutionalization. In short, comparative ethnography provides analytical leverage to examine how shared beliefs in familism as a strategy of action may lead to different decisional outcomes under varied circumstances of disease manifestation, structural position, and institutional circumstances.

Our comparative ethnographic analysis highlights the continuity and heterogeneity of decision-making experiences in Latinx households where individuals with cognitive decline remained at home (with or without support), transitioned to institutional care, or were navigating that transition. We examine longitudinal qualitative data to trace the temporal trajectories and deliberative processes of institutionalization decisions. Our analysis builds on existing ethnographic models of decision-making (Boholm et al., 2013) in which decision-making is a socially situated and culturally embedded process that unfolds over time and is shaped by horizons of meaning. We extend the framework to explicitly integrate how individual experiences are shaped by interactions across interpersonal, institutional, community, and policy levels, adding contextual components of the social-ecological model applied to dementia care (Bernstein Sideman et al., 2022). Combining both frameworks enable us to consider not only individuals’ choices or stated preferences from surveys but also to trace how decisional preferences and outcomes emerge through the wrestling of cultural expectations, material circumstances, support network availability, disease progression, and people’s capacity to navigate care systems. Figure 1 visually represents this integrated model.

Figure 1.

Decision-making analytic model based on Boholm et al. (2013) ethnographic approach to decision-making and Bernstein Sideman et al. (2022) Social Ecological Model of Health in Dementia Care.

Throughout this paper, we use cognitive decline to refer to a progressive process that may—but does not always—lead to a formal dementia diagnosis. We adopt this framing to better reflect the experiences of many Latinx families, for whom dementia is often diagnosed late, if at all (Hinton et al., 2024; Tsoy et al., 2021). Structural barriers—including limited access to care, linguistic and cultural mismatches in service provision, and provider bias—contribute to these delays (Fernandez Cajavilca et al., 2025; Quiroz et al., 2022; Vila-Castelar et al., 2022). Additionally, culturally rooted understandings of aging and memory loss may influence how symptoms are interpreted by older adults and their relatives and whether medical care is sought in the first place (Light et al., 2023, 2024). For some families, disclosing or naming a dementia diagnosis is stigmatized and shapes whether a diagnosis is accepted or pursued (Rodriguez & Padilla-Martínez, 2020). As a result, decisions about institutional care may begin in the absence of a formal diagnosis but in the presence of clear cognitive and functional impairments. While the concept of cognitive impairment is typically anchored to clinical assessments of cognition, we emphasize how it is evaluated and interpreted in family contexts—relationally, informally, and sometimes independently of biomedical timelines. By focusing on progressive cognitive decline, we aim to capture a broader set of experiences that shape care transitions, without artificially separating those who meet diagnostic thresholds from those who do not.

Research design and methods

This article is part of a larger longitudinal community-based study that employs comparative ethnography to explore the everyday experiences of older adults living with cognitive decline and their care partners in diverse communities in the US. During three years of community-based fieldwork, the study team explored and documented day-to-day experiences of people living with cognitive decline or caring for them. For the current analysis, ethnographic data were collected by a team of bicultural researchers trained in qualitative methods in community settings and institutional care facilities across the San Francisco Bay Area and Central Valley in California. Participants were recruited through purposive and network-based sampling in community-based settings such as senior centers, older adult day programs, and institutional care residential facilities. The sample size was not predetermined; instead, data collection continued until analytical saturation was reached. To ensure trustworthiness, the study relied on long-term participant observation, repeat interviews, reflexive memo writing seeking alternative explanations and disconfirming evidence, team-based iterative coding of qualitative data, and triangulation across fieldnotes and interviews. These practices reflect established standards for rigor in ethnographic research (Small & Calarco, 2022).

Eligibility for the parent study included (1) being an older adult expressing cognitive decline or (2) identifying as a care partner of someone living with cognitive decline. Participants were required to be conversant in English or Spanish, and capacity to provide informed consent was assessed prior to enrollment. For the current analysis, eligibility was limited to participants of the parent study who identified themselves as Latinx. Follow-up interviews were conducted approximately six months to one year after the initial interview, contingent on participants’ ability to participate. For older adults, this required sufficient cognitive capacity; for care partners, it required that the older adult was still living. In all cases, follow-up participation was voluntary.

To ensure consistency in data collection across researchers, all interviews were conducted using a shared semi-structured guide. In addition, interviewers completed a standardized training developed in collaboration with the UCSF Kanbar Center, which used a Standardized Research Participant approach—adapted from clinical simulation methods—to align interviewing technique and support reflexive, high-quality data collection (Arteaga, 2025). Semi-structured interviews were guided by an open-ended protocol that covered participants’ background, experiences with cognitive decline, use of formal and informal support services, and key decisions in their caregiving trajectory. The baseline and follow-up interview guides were used flexibly to follow participants’ narratives and allow for conversational depth in accordance with qualitative standards (Small and Calarco, 2022). They are included in the manuscript Supplementary Material as well as information about how our qualitative database was constructed to aid the retrieval of relevant ethnographic material for analysis.

This analysis—and the larger study of which it is a part—use comparative ethnography in the tradition of pragmatic realism, an approach that falls within a post-positivist paradigm (Lofland, 1995; Rendle et al., 2019). Pragmatic realism operates under the assumption social life is best understood and explained by analyzing interactions and interpreting meanings, it seeks to enhance scientific rigor and reproducibility through comparative designs and iterative analyses, and it values theory that helps advance our understanding of everyday life (Reeves et al., 2008).

Data included unstructured observations and interviews analyzed using Atlas.ti to support both deductive and inductive coding (Deterding & Waters, 2021; Lichtenstein & Rucks-Ahidiana, 2023). Deductively, our analysis built on the parent project’s attention to how structural conditions, institutional contexts, and cultural practices shape experiences of aging and cognitive decline. Inductively, we generated detailed memos based on fieldnotes and interviews with Latinx households, placing emergent insights in dialogue with our theoretical model to refine conceptual categories and significant relationships between them (Dohan et al., 2016; Emerson et al., 1995). Memos were drafted by the first author (I.A.) and reviewed by the second author (A.H.), a bilingual and bicultural team member with longstanding relationships in the study communities, who verified and supplemented each case.

In a second phase of analysis, we used the household narratives to create an ethnoarray, a technique to visually display qualitative data within and across cases (Abramson & Dohan, 2015; Garrett et al., 2019). The ethnoarray in this paper represents households (as columns) and decisions, resources, and sociodemographic characteristics of the households (rows), based on all associated case information in our qualitative database (Abramson et al., 2018). Its inclusion serves to visualize, complement, and triangulate traditional qualitative analysis through a combination of analytical modes (Small and Calarco, 2022). The supplementary material offers details on clustering decisions and replication code using R (R Foundation for Statistical Computing, Vienna, Austria).

Study procedures were overseen and approved by the Institutional Review Board (IRB) of the Human Research Protection Program of the University of California, San Francisco. In accordance with the IRB protocol, participants provided oral consent for observation and written consent for interview. The protocols included processes to ensure that all participants, including those experiencing cognitive decline, could provide meaningful informed consent. All data presented in this paper have been anonymized to protect confidentiality, but aside from anonymization, they have not been altered or manipulated.

Results

Twenty-four research participants from 22 households identified as Latinx in our larger comparative study. This sample included 10 care partners, 2 professional caregivers, and 12 older adults (60 years or older). This analysis excludes data from the professional caregivers and 3 older adults who were not experiencing cognitive decline. Our final analytic sample included 10 care partners of people with dementia and nine older adults living with cognitive decline. See Table 1 for details.

Table 1.

Sociodemographic characteristics of research participants

Ethnoarray ID	Pseudonyms	Gender	Caregiving relationship	Region (CA)	Age	Educational attainment	Generational status
Caregiver 09	Alejandro Morales	Male	Looking after sister	Central Valley	60–70	Less than high school	First generation
Caregiver 13	Daniela Torres (ab)	Female	Looking after husband	Bay Area	80–90	High school degree or equivalent	First generation
Caregiver 14	Sofia Torres (ab)	Female	Looking after father	Bay Area	50–60	College degree or equivalent	Second generation
Caregiver 24	Carlos Martinez	Male	Looking after wife	Central Valley	80–90	College degree or equivalent	Second generation
Caregiver 29	Gabriel Sanchez (a)	Male	Looking after wife	Central Valley	70–80	College degree or equivalent	Second generation
Caregiver 31	Lucia Fernandez (a)	Female	Looking after mother	Bay Area	70–80	Less than high school	First generation
Caregiver 39	Isabel Rivera	Female	Looking after husband	Central Valley	60–70	College degree or equivalent	First generation
Caregiver 42	Mariana Lopez (a)	Female	Looking after extended family	Bay Area	50–60	High school degree or equivalent	First generation
Caregiver 46	Valeria Gonzalez (a)	Female	Looking after husband	Central Valley	60–70	Some college education	First generation
Caregiver 49	Ana Ramirez (a)	Female	Looking after mother	Bay Area	40–50	College degree or equivalent	Second generation
Older adult 03	Elena Castillo	Female	Older adult in institutional care	Bay Area	70–80	High school degree or equivalent	Second generation
Older adult 07	Mateo Torres (b)	Male	Older adult passed in institutional care	Bay Area	90–100	Some college education	First generation
Older adult 13	Carmen Diaz	Female	Older adult moved into institutional care	Bay Area	90–100	Some college education	First generation
Older adult 14	Alicia Morales	Female	Older adult living at home	Bay Area	80–90	Some college education	First generation
Older adult 16	Hernan Vega	Male	Older adult living at home	Bay Area	80–90	High school degree or equivalent	First generation
Older adult 17	Teresa Flores	Female	Older adult living at home	Bay Area	70–80	Less than high school	First generation
Older adult 20	Luis Herrera	Male	Older adult living at home	Bay Area	70–80	Less than high school	First generation
Older adult 34	Rosa Medina	Female	Older adult living at home	Bay Area	70–80	Less than high school	First generation
Older adult 39	Claudia Navarro	Female	Older adult in institutional care	Central Valley	60–70	College degree or equivalent	Second generation

Table 1 summarizes participant characteristics, including study identification (ID), pseudonym, caregiving relationship, geographic area of residence within California (CA), age range (in years), educational attainment (by level), cultural heritage (self-identified), and generational status in the US (a) denotes bereaved care partners, and (b) indicates participants from the same household.

Among the care partners, five were bereaving a loved one who lived with cognitive decline, one had moved their loved one to institutional care during the study period, and three were considering relocation to institutional care. Among the older adults living with cognitive decline, four were already living in institutional care, while five were living at home. This reflects a diverse range of caregiving experiences and decision-making processes.

Figure 2 presents an ethnoarray decisions (purple), resources (yellow), and socio-demographic characteristics (green) across all 17 households in our sample (19 participants). Definitions for each element of the array appear in Table 2. Each row in the ethnoarray represents a theme that corresponds to observable patterns in participants’ narratives. Included rows visualize the factors most salient in our analysis. Shaded cells within the grid indicate the presence of a specific theme in a particular household while unshaded cells indicate the absence of that theme. Clusters represent groups of columns identified as sharing a pattern of similarity. The arrangement of these clusters is visualized with bounding boxes to indicate the limits of each identified group.

Figure 2.

Characterization of Latinx households in the study: Caregiving decisions, Resources, and Socio-demographics. This figure presents an ethnoarray that maps decisions (purple cells), resources (yellow cells), and socio-demographic characteristics (green cells) across 17 Latinx households (19 participants). Each column corresponds to a household; each row represents a theme. Shaded cells indicate the presence of a given theme, as defined in Table 2. Clusters of households are identified by high-level dendograms placed at the top of the grip. One household (Older adult 13) exists in its own category as an outlying case, which offers useful comparison.

Table 2.

Definition and operationalization of ethnoarray themes

Theme	Definition	Operationalization
Moved to Institutional Care	The person with cognitive decline transitioned to institutional care (e.g., Skilled Nursing Facility, Assisted Living Facility, Memory Care Unit, Continuing Care Retirement Community, or Residential Care Home).	Yes = 1; No = 0
Mention of Seeking a Diagnosis	Participant mentioned seeking a medical opinion to address concerns about cognitive decline, whether related to symptoms, diagnostic process, or care recommendations for themselves or a loved one with cognitive decline.	Yes = 1; No = 0
Private Financial Resources	Participant reports having income derived from personal or family wealth, retirement funds, pensions, or similar sources, enabling them to meet daily living expenses without relying on public assistance programs, such as disability income or means-tested government aid.	Yes = 1; No = 0
Medical Insurance Coverage	Participant reports whether they have active health insurance to cover medical or care-related expenses. Coverage includes Medicare but does not include California’s Medicaid program (Medi-Cal) unless participant is eligible for both programs.	Yes = 1; No = 0
Property Ownership	Participant reports that they or their spouse owns real estate or property, potentially contributing to financial stability in caregiving decisions.	Yes = 1; No = 0
Support Network	Participant reports they feel they have reliable personal connections available to provide assistance or emotional support in managing cognitive decline of themselves or a loved one.	Yes = 1; No = 0
History of Full Employment	Participant reports having a consistent work history, providing financial resources and work-related benefits such as a retirement fund.	Yes = 1; No = 0
Second Generation in the USA	Participant reports they were born in the United States to parents who emigrated from Latin America, as opposed to being born in Latin America and later migrating to the United States.	Yes = 1; No = 0

The ethnoarray illustrates the diversity of care decisions among Latinx households in Northern California, with five households featuring an older adult already living in institutional care (top row, purple squares). Turning to socio-demographics, larger proportions of households with second generation participants (5 of 6 households that include US-born children of immigrants) sought a formal diagnosis—a decision-making resource that legitimizes and facilitates access to services—than first generation households (7 of 11); reported movement of older adults with cognitive decline to institutional care (3 of 5 households) as opposed to first-generation households (2 of 11). Material advantages such as health insurance, financial assets, or homeownership are also seen in households on the right side of the array (from Caregiver 49, Ana Ramirez, onward), where caregiving transitions were more frequent. However, these groupings coexisted with variability in movement to institutional care, illustrating that decisions were not determined by structural resources or socio-demographics among participants in this study.

Our data also show that while there are substantial similarities within the sample, there is observable heterogeneity in structural position. High-level dendrograms on top of the grid reveal three main clusters of participants while further underscoring that decision making about institutional care can be explained neither by resources nor by acculturation.

To show how the array helps triangulate data sources in the ethnoarray, we examine the experiences of Mateo Torres (Older adult 07), a Nicaraguan-born man living with frontotemporal dementia. A long-time California resident with a history of stable employment, Mateo moved into institutional care when behavioral symptoms became unmanageable. One of his three US-born children—a college graduate and homeowner—was closely involved in the transition. This is represented in the shading in the rows for resources, exemplifying how the ethnoarray can visually synthesize conceptually informed qualitative data, offering a holistic view of how social positioning, resource access, and caregiving trajectories intersect.

These visual patterns support and extend our qualitative findings, while also pointing to disconfirming examples. Exceptions—like Older adult 13, Carmen Diaz, who relocated to institutional care without the typical indicators of structural advantage—underscore the limits of a visual summary and the rewarding insights from deeper ethnographic engagement for understanding the lived context of decision-making. As in Mateo Torres’ case, the columns represented in the ethnoarray were drawn from detailed analytical memos along with a reading of all text associated with the participant’s case. These memos help contextualize participants’ current care arrangements and anticipated needs. For instance, some caregivers (e.g., Caregiver 39, Isabel Rivera) described institutional care as a potential future option, but not yet necessary. Others (e.g., Caregiver 24: Carlos Martinez and Caregiver 29: Gabriel ­Sanchez) had reorganized household roles and secured paid caregiving to accommodate the progression of cognitive and functional decline. These cases illustrate the progressive nature of cognitive decline and the evolving decision-making it entails. The remainder of our analysis explores these complexities by providing a granular view of caregiving decisions regarding institutional care.

Trajectories to institutional care among Latinx households

Our findings show that Latinx households deliberate over time about institutional care, as older adults with progressive cognitive decline and their care partners weigh loved ones’ preferences against cultural expectations, available support networks, and the financial and bureaucratic constraints of a complex care system. We include three memos that illustrate the deliberative process.

The pace of change

In the Latinx families we came to know, caregiving needs typically escalated over time, often outpacing the family’s ability to secure additional unpaid or paid assistance. Household members had to continuously evaluate whether home care remained sustainable. For some, this led to a long period of consideration. Carlos Martinez’s story illuminates the temporality of making decisions about institutional care.

Carlos is an 86-year-old retired schoolteacher of Mexican background living with his wife in the California Central Valley. They have been together for 60 years, aging in the same town where they first met. As his wife’s dementia progressed, Carlos became her primary care partner, with emotional support from children living nearby and a professional caregiver supporting his wife’s hygiene and socialization needs for 10 hours a week.

Carlos finds himself in a decision-making dilemma as he weighs care options for himself and his wife, as her cognitive decline worsens and as incontinence issues ramp up. This makes him feel the burden of making decisions. Carlos and his wife enrolled in long-term care insurance once they managed to obtain an expert neurological assessment. The insurance now allows for financial reimbursement for caregiving services, so as a household, they would be able to cover their home care expenses and potentially able to afford moving into institutional care as a couple. Yet, despite this being financially possible, Carlos struggles with the idea of leaving her alone in an institutional care facility, because it feels to him like “putting her away.” He has leaned toward relocating together to be able to watch over her care. This would require making decisions about their current property and how to distribute the proceedings fairly among their children. He explains to the researcher how he feels about making this decision:

Carlos: I’m wrestling with that. We’re considering it. We’ve been considering it for the last 6, 7 months, but and my kids are, they all said, "Well, whatever you decide, Dad." And basically, the decision is mine. One of the factors that inhibits the speed at which I make the decision is that I don’t know how much time she has, and my health is getting worse all the time.

Carlos’s declining mobility and need for knee surgery are often neglected to keep taking care of his wife, but he is aware that at some point, it won’t be possible to keep her at home anymore, and she may need to move out with or without him.

Navigating financial and bureaucratic issues

For households that prefer institutional care, prohibitive cost can be a barrier to moving. Even after a decision is reached, moving into institutional care can present significant logistical and bureaucratic challenges. The Torres family’s experiences highlight the effort required to overcome this common challenge.

Daniela Torres, originally from Nicaragua, is the wife of Mateo Torres, who was once an electrical engineer and college instructor. The two met when as youth in Nicaragua but married much later when they reunited in the US. They raised three kids: Sofia, the youngest, and two older ones who died as adults. Mateo’s dementia diagnosis came after a stroke he suffered five years ago. Behavioral changes continued increasing until at some point, Daniela could no longer look after him at home. He was frequently insulting her, forgetting things, misplacing items, and getting lost in the neighborhood. His driver’s license was taken away by one of their daughters, a social worker.

A couple of years ago, the first daughter, who was helping her dad navigate the system passed away. Their daughter Sofia—a program manager, wife and mom of one—then stepped in. Sofia described the process of her father’s diagnosis, a step needed to tailor required supports in an institutional care facility, as “painfully slow,” with many missteps along the way. She recalls the diagnosis ordeal as a “big hot mess,” filled with fear and uncertainty. The bureaucratic hurdles that followed were immense. It took two years to establish conservatorship over her father to be able to take over his finances and pay for institutional care—nobody anticipated that the first two children of the marriage would die before their dad. Mateo had Medicare, some union insurance, and a small pension, but the healthcare system left the family to figure out much of the support on their own, a process that Sofia still resents. A researcher asked Sofia if she or her family had any support. She responded:

Sofia: This the US. We don’t get anything. No. No. No support, no nothing unless you have money and are very wealthy, yes, you’re pretty much screwed—or very, very poor. Very poor. Then you can get Medi-Cal, but then you have to live in a really, really awful places where it’s just—yeah. I'm sure you’ve seen those nursing homes.

To cover the cost of Mateo’s institutional care, the family decided to remortgage their parents’ home, which Sofia bought to live in, but which now needs major repairs. Sofia’s mom, Daniela, moved into the same institutional care facility to stay close to her husband during the pandemic, benefiting from a discounted rate. She still attends church regularly and brings communion back for her husband, who remains in the facility. The church community remains an important source of support.

Mateo passed away at the age of 101 in 2022, a few months after our initial interview. Though Daniela could have continued living at the facility, she chose to move out due to the cost. She stayed with Sofia for a few months before opting to live with a cousin, not wanting to burden her now only child.

The importance of pragmatism in decision making

For some participants such as Ana Ramirez, decisions about institutional care involved not only financial bureaucratic hurdles such as those the Torres’ family navigated but also required balancing pragmatic and family considerations. This is the memo with the story of her experience.

Ana is a second-generation Latinx woman from San Francisco, working as a teacher after studying marketing in college. She became the primary caregiver for her mother after a diagnosis of cardiovascular dementia, while her mom’s brother (Ana’s uncle) became the conservator of her mom’s assets accrued through lifelong real-estate investments, savings, and work pension. Ana’s mom initially lived in her own home, supported by Mexican caregivers twenty-four hours a day, as mom did not want to leave a house in which she had lived for the last 40 years. However, as in-home care became financially prohibitive and practically unsustainable, Ana made the decision to move her mother through various institutional care facilities over eight years, responding to changing medical needs and emergencies. The first move was to assisted living in the Bay Area, followed by a rehab center in San Francisco after a stroke, and finally to a small, licensed care home after seizures. Ana’s caregiving approach combined her belief in pragmatism and traditional expectations. Comparing herself to her cousins who also looked after an older adult with dementia in their family, Ana told a researcher: “You should not kill yourself trying to keep your demented mom at home as she wanted. You need to find a different solution”.

Ana chose private institutional care over lower-cost, public-funded options, which she described as unfit for her mother. “A madhouse,” she said, recalling a public facility with shared rooms, little to no support to residents, and poor infrastructure. Despite all her efforts, Ana voiced frustrations. She often felt that her mother’s needs were not being met promptly and that the palliative care services were insufficient. She advocated persistently for her mother’s care, visiting her several times a week and liaising with doctors and nurses to meet her mother’s evolving medical needs.

Discussion and implications: Heterogeneity matters

These ethnographic results highlight shared challenges and diverse experiences of Latinx respondents in their decisional process about institutional care. The findings demonstrate how decision-making is simultaneously informed by personal preferences, temporal considerations, resource constraints, and cultural orientations. The ethnoarray and the analytical memos illustrate that Latinx residents in Northern California urban and rural areas follow different trajectories when deliberating about institutional care—the group is far from homogenous. Our analysis included participants from a range of Latin American backgrounds, extending the literature’s focus on the Mexican American experience in California. The granularity of the analysis also adds to the range and complexity of socio-economic statuses (SES) represented in the literature. Epidemiological analyses often note the average lower SES of California-residing Latinx, but fine-grained ethnographic analysis of household experiences reveals the detailed experiences underlying demographic trends and data aggregation. Many participants in our study engaged with institutional decision-making while partnered, and our results thus shed light on enactments of familism within marital partnerships as well as between parents, children, and other family members.

Our data reveal that the decision to seek institutional care cannot be interpreted as an indicator of inadequate family regard, reliance, or support. Moving a loved one to institutional care does not mean lack of family commitment. As Ana and Daniela’s experiences illustrate, institutional care can include family and sustain familism; in both cases the family sustained their participation in care after the older adult moved to an institutional setting. In our data, several families considered moving into institutional care, such as Carlos Martinez, in order to “watch over” loved one’s care. These findings challenge either-or interpretations of family care versus institutional care. It is reasonable for quantitative analysts to interpret lower rates of institutionalization among Hispanic and Latinx older adults with ADRD as an indicator of a cultural desire to maintain a loved one at home (Herrera et al., 2008; Hinton et al., 2006; Light et al., 2023). But ethnographic data indicate that other social mechanisms—not just cultural preference—should also be explored to explain these patterns in quantitative data. Our findings suggest that familism can manifest in diverse ways within Latinx communities before, during, and after institutionalization.

The temporal length of decision-making, a recurring theme in participants’ experiences, also provides insights into the role of culture in the transition to institutional care. For participants such as Martinez’ family, deliberation was a long-drawn process that involved reconciling values, intentions, and socially ingrained expectations with realizations of available residential options. Participants weighed and explored options and preferences; they could sometimes control the process by creating new possibilities for support, such as hiring caregiving aides. Carlos’ experiences highlight how decision-making processes turned on self-assessment of caregivers’ physical and emotional limits as the toll of caregiving mounted on an aging body that grew weary responding to his wife’s steadily increasing needs. Temporality could also occur suddenly and unexpectedly as it did in Carmen’s case—an exceptional example shown in the ethnoarray—when a sudden fall meant she immediately required a level of care that her son, who was working long hours every day, could not offer. Anticipating and responding to care needs or drawing on network support could help families anticipate and navigate the need for institutional care (Prickett & Angel, 2017; Tate et al., 2023). Yet ethnographic data show that transitions in care could be unpredictable and difficult to control even for caregivers who were thoughtful and resourceful planners. In Latinx families, the dynamics of migration and transnationalism lent a degree of unpredictability to support networks (Angel et al., 2000). More importantly, cognitive decline and the transition to institutional care tend to defy rational planning and expectations (Lam et al., 2025). Cultural values did not directly guide caregivers as they made decisions about how to support family members but rather provided a set of tools that caregivers drew upon to make sense of and navigate the long temporal process of institutional deliberation.

Some participants received public assistance for in-home support for themselves or their loved ones. This federal program aims to facilitate aging in the community, but our participants’ experiences illustrate challenges, dependencies, and limitations of this support. Care partners faced bureaucratic hurdles, including liaising with long-term care insurance providers, settling household financial affairs, and devoting substantial time and effort to seek not only financial support but also a medical diagnosis. The experiences of Carlos Martinez, whose wife was living with advanced ADRD at home, illustrate that diagnosis is both unremarkable and essential to qualify for long-term insurance to enable their relocation to institutional care as a couple. Sofia—Mateo Torres’ daughter, described the diagnostic process as “painfully slow,” full of missteps and filled with fear and uncertainty. Our analyses thus demonstrate that obtaining a diagnosis was a critical milestone for many participants, yet the process felt protracted, fragmented, preordained, and ceremonial—a task of clearing bureaucratic hurdles rather than a moment of medical insight or revelation. Understanding the experience of navigating caregiving decisions—including how this navigation proceeds for those with and those without financial and social resources—may further elucidate the barriers Latinx families face in seeking a diagnosis and the role of this mundane yet essential medical fact in negotiating access to institutional care and other services (Hinton et al., 2006).

Even having cleared hurdles of diagnosis and eligibility, long-term care generated significant out-of-pocket costs and created financial strain for many families—particularly those who did not qualify for public assistance programs such as MediCal, California’s version of Medicaid (Mage et al., 2024). In the US, access to social support hinges upon means-tested eligibility assessments, and Latinx workers in low-wage jobs could fall into limbo—neither poor enough to qualify for MediCal nor prosperous enough to be able to afford adequate insurance (Herrera et al., 2008). Moreover, public programs to support older adults at home provided no guarantee of long-term stability. Eligibility for these in-home services turned on legal status, income, and assessed needs, and programs covered a limited number of professional caregiving hours each week. These restrictions meant that in-home services often proved insufficient as care needs of participants living with cognitive decline increased over time.

The heterogeneity within the Latinx population reflected in our findings also underscores the advantages of a sociological model of culture as compared to a frame of immigrant acculturation. Herrera, Angel and colleagues note that “time-honored cultural traditions and expectations of caring for older parents at home as opposed to placing them in a nursing home are at a crossroads.” (Herrera et al., 2012, p.260). Changes in caregiving practices may reflect shifts in structural and institutional circumstances or the changing individual needs of an older adult family member. An acculturation framework would associate keeping a loved one home as part of “Mexican” culture and institutionalizing a loved one as closer to “American” culture. In our ethnographic data, however, Latinx families who cared for a loved one at home, who placed them in institutional care, or who joined them in institutional care could all reconcile their actions as consistent with a strategy of familism. Practice models that emphasize how culture informs broad strategies of behavior (rather than particular choices) provide a more fruitful approach for understanding how Latinx families navigate care for older adults experiencing cognitive decline (Bourdieu, 1977; Freidenberg, 2000; Swidler, 2023) The cultural explanation of practice models adds to common explanations of structural vulnerability—which is experienced by Latinx individuals in California and in at least half of our participants—in assessing people’s available options when navigating care (Bourgois et al., 2017). Ana’s mother, for example, had the economic means to afford round-the-clock caregivers at home or institutional care. Her cousins, of lesser means, provided caregiving for their mother at home. Ana felt each practice could be reconciled with family values because, she noted, “you should not kill yourself trying to keep your demented mom at home as she wanted. You need to find a different solution.” Ana’s interpretation illustrates how cultural values, when strategically invoked and practiced, can inform a range of decisional outcomes.

These findings have important policy implications. In our data, cultural orientations such as familism and religion remain important in shaping Latinx families’ caregiving practices, but they did not have the automatic consequence of keeping loved ones at home. In the lives of our respondents, familial orientations were also consistent with the pursuit of institutional care. But many barriers remain; familial orientations do not dissolve the mean-testing accessibility requirement, and the assumption that all Latinx families desired home care for older adults with ADRD denies agency and generates an additional group-specific form of stigma. Health policy that recognizes the multiple ways in which institutional care is inaccessible—including by addressing potential stigma of the appropriate choice of institutional care—would constitute an improved family-oriented policy. This could have direct implications for the design and implementation of community-based strategies for dementia care outreach and service navigation, as well as changes in how primary care clinicians assess and respond to cognitive decline among their older adult patients of Latin American background.

This study has limitations. Longitudinal ethnographic field observations provide an unprecedented level of depth in charting responses to challenges in later life; however, the demands of the method necessitate geographical bounding (here, to Northern California) which may limit generalizability. Future work can benefit from replication to assess generalizability. Replication is especially important because structural and social determinants of health—including access to public benefits, housing, and caregiving infrastructure—vary significantly across the US and among Latinx subgroups, which may influence care pathways in other regions. The sample includes primarily individuals in the moderate to advanced stages of cognitive decline or their care partners. As a result, experiences of institutional care among those diagnosed earlier, or those more proactively navigating long-term planning, may differ. While this study does not aim for statistical generalizability, the patterns identified may be transferable to similar settings or inform future comparative research that considers institutional care decisions more broadly across Latinx populations.

Lastly, a note about discourse. In discussions about dementia care, institutional care is often framed negatively, seen in terms of a risk, a phenomenon that we should better predict, and that we should be able to mitigate. The connotation of that framing implies that institutional care is an undesirable public health outcome. This narrative can be harmful: it may stigmatize facility-based care, portraying it as a failure of individuals, care partners, or their support networks. Given the high costs, workforce challenges, and infrastructure limitations associated with aging populations, urgent improvement of the landscape of institutional care options is warranted to make high-quality and affordable care a priority (Angel et al., 2022; Reyes and Patterson, 2025). Still, respecting diverse care choices may help us accept diverse lifeworlds at a time when the promotion of diversity is under threat. Reducing the stigma attached to seeking support from institutional care facilities can alleviate the expressed guilt felt by caregivers facing challenging circumstances reported in the literature (Teng et al., 2020), resist further social disenfranchisement of people living with dementia (Beard & Fox, 2008), and promote informed, compassionate decision-making in dementia care overall.

Supplementary material

Supplementary material is available at The Gerontologist online.

Funding

This research was supported by the National Institute on Aging of the National Institutes of Health (NIA/NIH) award DP1AG069809 (Dohan PI). Content and views are those of the authors, not of NIH.

Conflict of interest

None declared.

Data availability

The code to construct the ethnoarray in R is available in the manuscript Supplementary Material to facilitate replication efforts. Due to confidentiality and ethical considerations, interview transcripts and fieldnotes from participant observations cannot be shared with other researchers. Access to these materials is restricted to protect the privacy of study participants, in accordance with ethical guidelines and institutional review board (IRB) requirements. This study was not pre-registered.

Ethical review

This study was reviewed and approved by IRB on 08/06/2020. IRB # 20-31940 Reference # 290813.

Informed consent 

Study participants gave oral and written informed consent following IRB protocols.

Study registration 

This study was not pre-registered, as it is a qualitative research study that did not require prospective registration.