This study used a qualitative descriptive design. Qualitative studies rely on text data and aim to understand the meaning of human action [16]. The current COVID-19–focused analysis is part of a broader study that planned to identify what ADRD care services are offered in primary care and how they are delivered across a variety of settings. This study was informed by the National Institute on Minority Health and Health Disparities Research Framework, which considers the complex and multifaceted nature of minority health and health disparities [17]. This framework includes different domains of influence (biological, behavioral, physical or built environment, sociocultural environment, and health care system), as well as different levels of influence (individual, interpersonal, community, and societal) within these domains.

The data collected were stripped of all identifying information and labeled with a code. The code can be linked to the individual’s identity, but the key linking the code to the identifying information is held separately from the research data. The University of Kansas Medical Center Institutional Review Board approved this project (STUDY00145615). Before the interview, all participants completed an informed written consent on the web either via their computers, tablets, or phones. The research team compensated participant’s time by posting them a US $40 gift card.

We recruited 2 groups of participants: (1) family caregivers of people living with ADRD and (2) PCPs. Inclusion criteria for caregivers included being aged 18 years or older, identifying as a close friend or relative of a Latino person with ADRD diagnosed by a health care provider or research study, providing or having recently provided care to them at least once a week in person or via phone, being proficient in English or Spanish, living in the United States, and being willing to participate in the study. Inclusion criteria for PCPs included being a medical doctor, doctor in osteopathic medicine, nurse practitioner, or physician assistant who currently or recently provided primary care services to Latino families with ADRD in the United States.

Caregivers were recruited via convenience sampling from diverse sources, including a research registry, clinics, and community ADRD patient lists, internet and newspaper advertisements, and connections with community partners and research team members. PCPs were initially recruited using snowball sampling techniques: first by contacting connections of the research team and later asking interviewed PCPs for referral of other PCPs.

We used purely qualitative semistructured interviews, which allow for the comparison between participants while allowing spontaneous exploration of topics relevant to unique participants [18]. The interviews took place between November 2020 and April 2021. All but 2 interviews were conducted via secure videoconference. The other 2 interviews were conducted over the phone. The interviewer used 2 audio recorders to reduce the risk of data collection failure and took notes at the end of each interview to summarize the findings. Questions were not provided to participants before the interview. No one else besides the interviewer and the participant was present during interviews, although in a few instances, participants had to pause the interview briefly to address issues with their loved ones (caregivers) or colleagues (PCPs).

The process for each interview was similar: JP-P interviewed all participants. The interview started with a short conversation aimed at developing rapport and explaining the main goals of the interview. The first questions asked about the basic characteristics of the participants. Core questions of the survey asked about participants’ experience with primary care clinics (caregivers) and serving Latino individuals with ADRD (PCPs). Unless participants had mentioned it spontaneously, the interviewer asked participants halfway through the interview how the COVID-19 pandemic had impacted them, their care recipient (caregivers), and the Latino families with ADRD they serve (PCPs; see Multimedia Appendix 1 for the interview guides). The interview questions had been previously pilot-tested within the team. The interviewer audiotaped all interviews, which were designed to last 45-60 minutes, including the COVID-19 questions. Interviews were conducted in English or Spanish, and those with caregivers included simple language to account for different levels of literacy. A professional team transcribed all interviews and the interviewer reviewed them for accuracy.

To bring rigor and validity to the research process, the interviewer used active listening techniques during the interview aimed at confirming the information shared by participants. The potentially lower theoretical ADRD expertise by family members and PCPs could create a perceived power differential with the interviewer. In fact, some family members and PCPs expressed worry about the interviewer potentially testing their ADRD knowledge. To reduce this perception of power differential, the interviewer emphasized that the interviews would ask about their experiences and that they were the experts in those experiences. JP-P already knew some of the family caregivers via his research but knew none of the PCPs. To balance how much participants knew about him, he introduced himself, his background, previous work, and research goals the first time he contacted them and at the beginning of the interview. Both coders (JP-P and MF-C) had previous coding experience. The fact that only 1 coder conducted the interviews and that they both had different backgrounds (JP-P: PhD, man, faculty, and psychology and public health education vs MF-C: MD, woman, research associate, and medical education) allowed different perspectives.

We organized the transcripts and notes for qualitative review, using a pragmatic approach and thematic analysis methods, which emphasize identifying, analyzing, and interpreting patterns of meaning within qualitative data [19,20]. We organized the data in Dedoose [21]. JP-P identified relevant sociodemographic, ADRD relationship, and clinical service data in the interviews and summarized it into descriptive statistics. JP-P and MF-C independently read the interviews and notes once initially to familiarize themselves with the data, coded the content of the text by identifying codes and themes, and resolved coding disagreements through discussion and consensus.

Considering the time-sensitive nature of this work, we decided to stop recruitment when data saturation was achieved at 44 participants. Table 1 shows the characteristics of the 21 family caregivers and 23 PCPs. Most caregivers were women (19/21, 90%), younger than 66 years (16/21, 76%), and children of the care recipient (15/21, 71%), most of whom had Alzheimer disease (14/21, 67%). All but 1 (20/21, 95%) participant identified as Latino of diverse origins; 6 (29%) were born in the United States; and all lived in urban regions in the Midwest (15/21, 71%), the Northeast (5/21, 24%), and the South (1/21, 5%). Interviews were conducted in Spanish with 13 (62%) caregivers, 14 (67%) of them reported good English proficiency, and 8 (38%) reported that their care recipient had good English proficiency.

More than half of PCP participants were women (13/23, 57%) and US born (12/23, 52%). A total of 10 (43%) PCPs identified as Latino of diverse origins and another 10 (43%) identified as non-Latino White. A total of 18 (78%) PCPs lived in urban settings, with the Midwest being the most common region (15/23, 65%). A total of 12 (52%) PCPs reported being able to communicate with their patients in Spanish. In all, 15 (65%) PCPs were medical doctors and 7 (30%) were nurse practitioners; the types of clinics they practiced in varied, along with a wide distribution of the percentage of Latino patients served.

We developed 8 themes from the analysis. Table 2 shows the themes and codes identified in the qualitative analyses.

The physical domains described by caregivers and PCPs included accelerated deterioration, poor nutrition, and death.

Even though participants understood the need for the lockdown, they frequently described how it contributed to social isolation and deprivation, which, in turn, reduced engagement in activities and led to depression and apathy.

The pandemic’s impact on Latino caregivers’ working situation included avoiding exposure to infection, disrupted work, and increased work burden.

The pandemic impacted multiple systems that support care recipients and caregivers, including health care and community care systems.

Families and PCPs highlighted the impact Latino families with ADRD experienced and the reduced access and quality of health care and community care services. In response, these services adapted to the COVID-19 pandemic context by leveraging home health and remote communication.

As mentioned earlier, remote communication became paramount during the COVID-19 pandemic for health care and community care services. Remote communication also became an important tool for families. However, Latino families with ADRD experienced barriers in communication related to their low trust, insufficient skills, diminished abilities, and inadequate support.

Caregivers were crucial in protecting care recipients’ health and maintaining their quality of life during the pandemic. In fact, caregiver support was the most frequently mentioned type of social support. Caregivers largely supported care recipients in protecting against COVID-19 infection and reducing the psychosocial, health care, and community care impact of the COVID-19 pandemic.

This study aimed to gain an in-depth understanding of the impact of the COVID-19 pandemic among Latino families with ADRD, from the perspectives of family caregivers and PCPs. To achieve this goal, we interviewed 21 family caregivers of Latino individuals with ADRD and 23 PCPs across the United States. These participants were diverse with respect to their region, primary language, and other characteristics. We found that the COVID-19 pandemic has impacted Latino families with ADRD at multiple levels, ranging from physical to health care and community care levels. Latino family caregivers’ actions were key to addressing these new challenges.

Some impacts of the COVID-19 pandemic on Latino families with ADRD resembled impacts reported in previous research for the general population. For example, accelerated cognitive deterioration was also reported in a mixed methods study in the United Kingdom where 184 caregivers and 24 people with ADRD were interviewed [22]. The lack of access to engaging activities and human contact were also reported in the UK study, which were found to be potential drivers of the accelerated deterioration. Other similar findings from this study include the impact on care recipients’ and caregivers’ mental health, the barriers with the technology needed to replace in-person contacts, and the importance of caregivers to bring about resilience despite the adversities. Similar impacts have also been reported among family caregivers of people with ADRD in India and Italy [23,24]. Despite these similarities, our findings highlight the need to pay attention to pandemic-related stress that add up with many other life stresses found frequently in minoritized populations [25], as well as additional factors that compound the already existing digital divide’s impact on remote communication among Latino individuals: low trust in these technologies and decline in technological abilities among people with ADRD [13,23,24].

Some impacts have been reported less frequently or not reported at all previously. To our knowledge, no study has reported the impact of the COVID-19 pandemic on nutrition among people with ADRD and their families. This impact is in line with a systematic review of 28 studies across the world, where most of them identified changes in food intake toward the adoption of unhealthy eating among people of all ages [26]. Our findings shed light on potential solutions for the Latino community, as those impacted tended to be male caregivers with low income and families for which the meal delivery services did not provide culturally appropriate meals.

We also found that the COVID-19 pandemic impacted caregivers’ work. This finding aligns with a study from India where many caregivers had to start working from home or faced disruptions that affected their finances [24]. In our Latino sample, the pandemic did not just disrupt work for some but also increased the work burden of others. Latino individuals may have been especially impacted by the work consequences of the pandemic, as they are overrepresented in the frontline workforce, tend to live in multigenerational homes, and have among the lowest average incomes in the United States [27-29].

Our findings align with 2 commonly reported factors among Latino individuals that have been shown to impact health behaviors, including health care use and assertiveness [30-32]. First, our interviews revealed that the news about Latino individuals’ disproportionate risk of COVID-19 infection and death [9] may have increased the fear among Latino families and contributed to their low use of sometimes critically needed resources. These findings align with “fatalism,” a belief that events are beyond one’s control [31]. Second, we also found that to reduce the risk of infection, PCPs reduced physical contact with their patients, which they saw as a barrier to care. This finding suggests an impact on “personalism,” or the importance of establishing relationships through warm interactions [31].

This study has some limitations. Remote recruitment and interviews increased the representation of participants in rural areas and other states. However, video calls and phone calls led to some communication issues, which, in some cases, reduced the amount of information we could collect and affected the quality of the audio. The inability to conduct in-person recruitment and interviews may have excluded the most underserved individuals, who could have been contacted via health fairs before the pandemic started. We did not interview individuals with ADRD, which did not allow a full triangulation between them, their caregivers, and PCPs. Although Latino caregivers tend to be women [33], they were overrepresented in our study, likely also due to women’s higher likelihood to participate in health-related research [34,35]. The sample size was relatively small and not probabilistic, which reduces the generalizability of the findings. We did not return the transcripts to participants for comment or correction. As with most studies, individuals who participated in the study were motivated to participate. We do not know how much their discourse compares to those who decided not to participate.

This study has implications for public health. Given the efficacy of existing COVID-19 vaccines [36], ensuring access to ongoing boosters among Latino individuals with ADRD and their families will be needed. To do so, it will continue to be necessary to hold events at flexible times and days; provide care at convenient venues; and improve the communication with Latino individuals by using a wide range of communication modalities (eg, calling, texting, and patient portals) in a way that is trustworthy, easy to use for people with cognitive impairment, and linguistically and culturally appropriate. The common stress and depression related to the fear of COVID-19 infection, uncertainty, and confinement among Latino families with ADRD highlight the need improve access to mental health services in general and specifically during pandemics. An example of a potentially inexpensive intervention is layperson-delivered, empathy-oriented telephone call programs, which have been shown to reduce loneliness, depression, and anxiety [37]. These services can be provided by governments or charity organizations and offered via primary care clinics, health departments, and entities that identify those who are potentially in need. Other useful services may include cognitive and physical engagement activities to reduce confinement-related deterioration. Since stresses are cumulative and Latino individuals tend to have higher levels of socioeconomically driven chronic stress [25], building stronger and more accessible social and health care welfare systems will also be key.

The impact of the pandemic on physical and cognitive decline, potential sources of support, and the crucial importance of families in the care of Latino individuals with ADRD highlight the need to provide them with financial support for their services, respite when possible, and accessible caregiving support training. Home care services have been essential in caring for some Latino individuals with ADRD and providing respite to their caregivers. However, these professions tend to be poorly paid, and their services are hard to access. Home care workers’ salary could be adjusted to the value of their service and covered by health insurance companies or government programs.

Our findings can inform future studies. Our findings regarding the physical and cognitive deterioration caused by the pandemic and the importance of family support may help inform future studies on the differential impact on ethnic groups in community and institutional settings and on the buffering effect of social support and health care and community care services. For example, nursing homes that have a higher proportion of ethnically minoritized populations had higher rates of COVID-19 infection and death [38]. Studies could explore whether this disproportionate impact also applies to cognitive and functional decline. Our data suggest that Latino families where men were the primary caregiver were at risk of poor nutrition and that home meal delivery services not offering culturally appropriate meals was a barrier to improving nutrition quality. Future studies can develop remote interventions to provide education on how to cook healthy, culturally appropriate meals for Latino family caregivers and explore the cost-effectiveness of offering Latin American healthy meals in communities where Latino individuals are present. Our findings also suggested that the awareness that Latino individuals had a higher infection and death risk from COVID-19 might have reduced their use of health care services. Future studies should develop culturally appropriate messaging in response to disparities where the fear and related factors facilitate health behaviors instead of reducing them.

In this study, we have found that the COVID-19 pandemic has impacted Latino families with ADRD beyond infection and physical symptoms, and family caregivers have been crucial to maintaining care recipients’ health and quality of life. The experiences of Latino families with ADRD during the pandemic resembled those of the general population. These experiences included a lack of access to engaging activities and human contact, issues with the technology needed to replace in-person contacts, accelerated cognitive deterioration, and an impact on caregivers’ and care recipients’ mental health. However, this pandemic has revealed many of the barriers that Latino families with ADRD face, and in most cases, this pandemic has exacerbated previous barriers. These barriers included nutrition being affected by inefficiencies of the mailing system or financial insecurity; finances being affected by the pandemics’ impact on jobs typically held by Latino caregivers; and fatalism and personalism interacting with Latino individuals’ disproportionate risk to COVID-19 infection and death, which reduced health care use. These findings suggest the need for more equitable access, culturally appropriate and trustworthy content and delivery of health care and community services, as well as stronger financial and social supports for family caregivers.