Evaluation of social determinants of health and emotional symptoms among rural family caregivers providing palliative and end-of-life care: An explanatory sequential mixed methods study protocol

Abstract

Introduction

Approximately 53 million family caregivers (FCGs) support and provide home-based care for patients with chronic and life-limiting conditions. Caregiving at the end of life is an emotionally, cognitively and physically demanding role, and FCGs are often unpaid and receive little training. FCGs in remote, rural areas experience adverse social determinants of health (SDOHs), such as limited access to healthcare services and are often financially vulnerable, affecting their healthcare coverage, placing them at higher risk for poor emotional health. Adequate health literacy, a key SDOH, may be a protective factor. The purpose is to present the protocol of an explanatory sequential mixed-method study aiming to examine rural-dwelling FCGs’ emotional health, relationships among FCG characteristics (demographics, co-morbidities) and SDOH and health literacy as a potential mediator.

Methods and analysis

With 80% power to detect a mediation effect of at least 0.11 when the two-sided alpha is 0.05, an anticipated 108 adult rural FCGs who are providing palliative end-of-life care (PEOLC) will be recruited. Rigorous remote, digital- and community-based recruitment strategies will be employed to increase access to FCGs from under-represented rural communities. Self-report surveys include demographic and health characteristics, SDOH (health literacy, education level, employment status, annual income level, healthcare access and insurance) and anxiety/depressive symptoms. Descriptive statistics will be used to characterise the sample. Correlation analysis and one-way Analysis of variance (ANOVA) will be used to study the relationships between FCG characteristics, SDOH and anxiety/depressive symptoms. Hayes’ PROCESS Macro for IBM SPSS Statistics will be used to test mediation effects of health literacy between FCG characteristics/SDOH and anxiety/depressive symptoms. Indirect effects will be tested using bootstrapped CIs. Semistructured interviews will be conducted with a representative sample to elicit FCG perceptions relative to how SDOH affect their care provision and emotional health, then analysed using content analysis. Quantitative and qualitative strands will be examined and compared for convergence and divergence, then integrated for final interpretations.

Ethics and dissemination

Michigan State University’s Institutional Review Board has reviewed and approved the study. Written informed consent will be obtained from each participant. Study findings will be disseminated through publications in peer-reviewed journals and presentations at academic conferences.

Conclusions

Study outcomes will inform the role of SDOHs, particularly health literacy, for the development of interventions to emotionally support rural FCGs.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Mixed methods design: The explanatory sequential mixed methods approach allows for both statistical analysis and rich qualitative insights, enhancing the depth and validity of findings.

• Targeted recruitment strategies: Use of both remote (eg, social media and ResearchMatch) and community-based recruitment increases accessibility and inclusivity for hard-to-reach rural populations.

• Mediation analysis: Use of Hayes’ PROCESS Macro and bootstrapping strengthens the rigour of statistical mediation testing, allowing for nuanced understanding of health literacy’s role.

• Cross-sectional design: The current study cannot assess changes over time or establish causality due to its cross-sectional design; future work with longitudinal data will inform these areas.

• Limited broadband access: While addressed, technical and geographical barriers may still affect participation for some rural caregivers.

Introduction

Nearly 70% of the 1.72 million hospice patients in the United States (U.S.) received home-based care.¹ These numbers are expected to greatly increase during the next decade as the US faces surging numbers of ageing adults with serious, life-limiting illnesses, such as cancer, who are living longer and desire to remain at home during their final days.² As home-based deaths align with patient preferences, there is a growing reliance on friend and family caregivers (FCGs) for the provision of palliative and end-of-life care (PEOLC).³ FCGs are instrumental in assisting their patient with complex symptom management, emotional support, advance care planning and surrogate decision-making.^{4 5} However, FCGs often have no prior experience and/or receive little formal education or skills training; consequently, they may feel underprepared and lack adequate information essential to support their caregiving.⁶ Compounding this burden, FCGs in rural areas are often at a greater disadvantage, given limited access to and availability of support and PEOLC services.⁷

Nearly 20% of American residents live in remote, rural areas, often geographically distant from healthcare services.⁸ This decreased access to supportive resources in rural areas exacerbates the shift of care responsibilities to the FCG, who provide an average of 26.3 hours/week of care.9,11 Additionally, compared with urban residents, those in rural areas are older (17.5% vs 13.8% ages >65 years) and in poorer health, with higher rates of co-morbidities (≥2 medical and/or psychological conditions).¹² This suggests that many FCGs are managing their own health conditions while concurrently overseeing the care for their care recipient. Furthermore, rural-dwelling FCGs have greater financial disadvantages and comparatively less formal education (19% vs 33% with >bachelor’s degree).¹² Adverse social determinants of health (SDOHs), which include lower health literacy, income, education, employment status and insurance status, are recognised to negatively impact rural-dwelling FCGs’ ability to provide adequate PEOLC and manage their own health.¹¹

Given FCGs’ personal characteristics and SDOH that increase vulnerability, coupled with the stressors associated with the provision of PEOLC for a loved one, rural FCGs often experience adverse emotional health.¹³ Moderate to severe emotional symptoms are common among general hospice FCGs, with approximately 33% experiencing anxiety and 25% experiencing depression.¹³ Current research on rural-dwelling hospice FCGs’ emotional symptoms has been mostly qualitative, demonstrating gaps in our understanding relative to relationships among FCG personal characteristics, SDOH and adverse emotional symptoms.¹⁴ Examining these relationships in more depth is essential for increasing insight into factors that heighten rural-dwelling FCGs’ vulnerability to adverse emotional health.

Caregiving heavily relies on accurate and timely information for providing effective, high-quality and safe care at home. FCGs who have informational deficits experience heightened perceived stress¹⁵; thus, adequate health literacy, a key SDOH, may be a protective factor for rural FCGs’ emotional health. Health literacy is an inherently individual and systemic process involving individuals’ knowledge, motivation and competencies to access, understand, appraise and apply information to make judgements and decisions in everyday life.¹⁶ Health literacy is necessary for FCGs to make decisions confidently, to seek healthcare when needed, and to manage personal emotional well-being.¹⁷ Limited FCG health literacy is associated with increased anxiety and depressive symptoms, secondary to stressors associated with greater unmet informational needs.^{11 18} Importantly, health literacy is a modifiable SDOH that influences health outcomes.¹⁹

An integrated conceptual framework developed by Sørensen and colleagues will be used for examining health literacy and other SDOH factors in the context of rural-dwelling FCGs and their PEOLC.¹⁶ Relevant to the current study, Sørensen and colleagues presented a central tenet, in which health literacy is influenced by antecedents, namely proximal factors like personal determinants (eg, age, race, gender) and distal factors (societal and environmental impacts).¹⁶ The model provides a foundation for better understanding the specific personal characteristics and associated SDOH that place rural FCGs at risk for inadequate health literacy, and therefore, the burden of poorer emotional health. Furthermore, the integrated model presents pathways that link health literacy to health outcomes, offering insight into underlying mechanisms for which these characteristics and SDOH impact FCG emotional health outcomes.

Previous studies propose health literacy as a potential mediator in the mechanisms by which certain SDOH (eg, education, employment status) affect emotional health outcomes.20,22 However, these studies used general population samples; whereas, rural communities are under-represented and underexamined. Rural-dwelling FCGs are a distinct group with unique conditions and needs.20,22 For example, a recent review found limited inquiry of other SDOH factors such as health literacy, food security, safe housing and nontoxic environments that may impact rural FCG emotional health outcomes.^{14 23} Exploring the mediation potential of health literacy could provide insight into how to best support FCGs as they manage the complexities of delivering PEOLC. Research is needed to better understand the rural caregiving experience and the SDOH that impact their emotional well-being and capacity to provide PEOLC, to improve current services and resources offered.

Purpose

This study proposes to investigate relationships among FCG personal characteristics (demographics, comorbidities), SDOH (education, income, employment, insurance status, health literacy) and emotional health (anxiety and/or depressive symptoms), and to evaluate the extent to which health literacy mediates these relationships among rural-dwelling FCGs providing PEOLC. The study will evaluate the following research questions:

1. What is the relationship between FCG personal characteristics and their emotional health (anxiety/depressive symptoms)?

2. What is the relationship between FCG SDOH and their emotional health (anxiety/depressive symptoms)?

3. Is there a mediating effect of health literacy on the relationship between FCG personal characteristics and anxiety/depressive symptoms?

4. Is there a mediating effect of health literacy on the relationship between FCG SDOH and anxiety/depressive symptoms?

5. What are FCG perceptions of SDOHs that impact their capacity to deliver care and their personal emotional well-being?

Methods

Design

The study uses a cross-sectional, explanatory, sequential mixed methods approach. A mixed methods design is suitable for this research as the concepts of interest among rural-dwelling FCGs are poorly understood and scarce, thus warranting both quantitative and qualitative inquiry that integrates findings. First, a quantitative component using self-report surveys will establish statistical relationships among FCG personal characteristics, SDOH including health literacy and emotional health, operationalised as anxiety/depressive symptoms. Second, a qualitative inquiry will evaluate perceptions of SDOH that impact care delivery and emotional well-being with a subset of the participants. Both data strands will be examined and compared for convergence and divergence.^{24 25} The data will then be integrated for final interpretations.^{24 25} The mixed methods design allows for data triangulation to corroborate and generate a comprehensive understanding of the proposed work.^{25 26}

Participants and setting

Power analysis

PASS 2020 was used to compute the required sample size. Using a test of mediation effects in linear regression and a 20% refusal rate as informed by past palliative FCG studies, a sample of 135 will be approached for enrolment.²⁷ An anticipated final sample of 108 achieves 80% power to detect a mediation effect of at least 0.11 when the two-sided alpha is 0.05. A sample size of 108 will allow us to obtain a robust estimate of the effect sizes for the study outcomes.

Eligibility criteria

The inclusion criteria will consist of (1) family/friends of an adult patient with a serious, life-limiting illness, meeting hospice and/or palliative eligibility; (2) adults, ages ≥18 years old; (3) self-identifying as providing care in a rural area and (4) ability to understand and speak English, as surveys and interviews are written at an English 5th grade reading level, measured by the Flesch-Kincaid test. FCG will be excluded if they provide care to paediatric patients, age <18 years.

Recruitment

The study is anticipated to be open for enrolment from March 2025 through January 2026. Convenience sampling will be used to recruit participants, incorporating various remote and in-person strategies. Remote strategies offer opportunities to reach rural-dwelling FCGs, who may have challenges related to travel, such as lengthy commutes and unpaved roads.28,30

Remote, digital recruitment strategies have shown promise in increasing study awareness and accessing hard-to-reach rural communities.³¹ Social media and other online methods facilitate more inclusive research participation. Paid advertisements on social media platforms, including Facebook, will target FCGs. Additionally, study fliers will be posted in community forums specific to residents in rural areas of the Midwestern state where the research will be conducted. Using an online registry of research volunteers on ResearchMatch, invitation emails will be sent to users who fit the eligibility criteria.

Notably, rural communities often have mistrust.^{14 32} Access and engagement with rural communities are facilitated by familiar and trusted sources within the community; therefore, in-person and community-based approaches increase the likelihood of effectiveness at reaching and engaging rural FCGs in research. Strong partnership with well-known community organisations offers a bridge between the rural community and the research team, facilitating study interest and recruitment.³³ There will be concerted efforts to increase study awareness by promoting the work in the spaces that FCGs frequent. For example, flyers will be posted on public bulletin boards within local churches, ageing agencies and senior centres in rural areas. Trusted members of these settings, often individuals who know the community well (eg, leaders of faith-based organisations and outreach directors), will serve as study advocates, encouraging FCGs to participate. In addition, through a collaborative effort with state-wide rural-serving hospice and palliative care agencies, fliers will be mailed to friends and family members of patients receiving services. Lastly, community outreach and engagement at events and groups, such as caregiver retreats, conferences and support groups, will be utilised. Using the phone numbers or emails listed on the study fliers, FCGs will be encouraged to contact the research office with questions and/or interest, along with their phone number and the best call-back time to discuss the study and complete screening.

Screening

Interested FCGs will be contacted by phone or email to review the informed consent detailing study information, including risks, benefits and rights as a volunteer participant. Within the form, there will be an option for FCGs to indicate their willingness to participate in a follow-up interview.

Following consent, FCGs with internet access will complete a short intake form involving a standardised screening survey. The survey will include questions related to inclusion/exclusion criteria to determine their eligibility. FCGs who are ineligible will be thanked for their time, while those who are eligible will be emailed a Qualtrics link with the study surveys.

For those who lack broadband, a copy of the informed consent and surveys will be mailed, along with a stamped return envelope addressed to the study office. FCGs will sign the consent document and complete the surveys before returning the forms to the study office. The study contact information, such as the phone number and email address, will be listed on the consent form for participants to contact for questions and/or study concerns. In addition, verbal consent and responses to surveys can be obtained through phone calls.

Data Collection

Quantitative procedure

Once enrolled, eligible participants completing surveys online will be sent a follow-up email with the questionnaires. The approximate survey completion time is 10–15 min, as pretested with older adults from the general population who served as mock participants. Research team members will be available by phone to answer questions regarding the consent form and surveys.

At the end of the Qualtrics survey, there will be an automatic message displayed with an embedded link optional for participants to follow, which will redirect them to a website with credible emotional health information (eg, the National Institute of Mental Health website).^{34 35} Information on anxiety and depressive symptoms and common management/treatment options will be included.^{34 35} If symptoms are present or FCGs have questions, the postsurvey message will encourage them to discuss with a healthcare professional. In addition, the message will include a link to the state’s Health and Human Services website for participants to visit and enter their postcode to locate nearby free or low-cost local community mental health resources.³⁶ Alternatively, resources will be mailed to those without broadband.

Questionnaires

Demographics/personal characteristics

A sociodemographic survey with single-item forced-choice questions for age, sex, race/ethnicity and the relationship to the patient will be administered. Co-morbidities, as a proxy to FCG health status, will be measured using the 21-item Bayliss Co-morbidities checklist, and will indicate the presence of common chronic conditions.³⁷ Compared with chart review, the checklist has a median sensitivity and specificity of >75% and 92%.³⁷ The checklist has been previously used in caregiver samples.^{38 39}

SDOH: health literacy

The Health Literacy Survey-EU (HLS-EU-Q16) is a 16-item self-report survey that assesses the perceived difficulty of four dimensions of health literacy related to accessing, understanding, appraising and applying healthcare information.⁴⁰ The HLS-EU-Q16 is conceptually grounded and multidimensional in content.^{16 41 42} Response options are on a 4-point scale, ranging from 1 (‘very difficult’) to 4 (‘very easy’).¹⁵ Higher scores represent better health literacy levels. When used among cancer FCGs, internal consistency was α=0.9.¹⁵

Other SDOH

A survey with single-item forced-choice questions will be administered to assess SDOH. Standardised measures from the NIH PhenX SDOH Assessments Collection will include questions on education level, employment status, annual income level and insurance status and type.⁴³

Anxiety and depressive symptoms

The PROMIS v1.9 Anxiety short form and PROMIS v.10 Depression 8a short form are both 8-item surveys.⁴⁴ Participants are asked to recall their symptoms (eg, ‘My worries overwhelmed me’) within a 7-day time frame. Items are rated on a 5-point scale that ranges from 1 (‘never’) to 5 being (‘always’). The internal consistency was α=0.93 when used in a sample of lung cancer caregivers.⁴⁵ Patient-Reported Outcomes Measurement Information System (PROMIS) instruments are often used in caregiving research, as they allow for use in both a clinical sample and a general population sample.⁴⁶

Interview procedure

Participants will be selected through nested sampling, using a subset of participants who completed the surveys, presenting with high scores on anxiety and depressive symptoms. Within 2–4 weeks of completing surveys, participants who indicated their willingness for follow-up interviews on the consent form will be contacted via email or telephone.

Based on the participant’s availability, the principal investigator (PI) will schedule an interview, inviting them to share their experiences and perceptions of providing PEOLC in rural areas within the context of SDOH, with an emphasis on health literacy skills and any SDOH with a strong relationship to emotional symptoms in the quantitative component. Moreover, FCGs will be asked to share any additional socio-geographical challenges that they experience, which may not have been captured quantitatively. Each interview will last approximately 45 min to 1 hour. Interviews will be conducted using Health Insurance Portability and Accountability Act (HIPAA)-compliant videoconferencing technology that enables computerised audio-video communications.⁴⁷ If technical issues arise, participants will be encouraged to call the study number for assistance. For interviews with participants without broadband, the dial-in feature will be used, which necessitates audio-only for a phone call interview.

Data management and analysis

SPSS (V.26) will be used for data management and statistical analysis. Descriptive statistics will be used to characterise the sample.

RQ 1, which examines the relationship between FCG characteristics and emotional symptoms, will use correlation analysis for the continuous variables (age, comorbidities) and a one-way ANOVA to study relationships between categorical FCG characteristics (sex, race/ethnicity, relationship type) and continuous anxiety/depression scores.

RQ 2, which examines the relationship between SDOH and emotional symptoms, will use a one-way ANOVA to study the relationship between categorical SDOH (education, employment, annual income, insurance status) and continuous anxiety/depression scores.

RQs 3 and 4 will examine the mediating effect of health literacy on the relationship between FCG characteristics and anxiety/depressive symptoms. The relationship between FCG SDOH and anxiety/depressive symptoms will use multiple regression to assess all variables in the mediation model. The PROCESS Macro for SPSS, a software package, will be used to follow the procedures outlined by Preacher and Hayes to test for mediation and moderation. For mediation analysis, a 2000-sample size bootstrap procedure will be used and specified at a 95% CIs to test for the significance of relationships. Indirect relationships will be considered significant if the CIs do not contain 0. After direct relationships are established, health literacy will be introduced to test its mediating effect on the relationship between (a) FCG characteristics and their anxiety/depression scores; and (b) SDOH and their anxiety/depressive symptom scores.

RQ 5 will examine FCG perceptions of SDOHs that impact their emotional symptoms, which will be analysed using content analysis.²⁶ Two researchers will independently review and code the transcripts and field notes. They will meet to discuss the coding until a consensus is achieved, thereby maximising the trustworthiness of the findings and establishing triangulation.²⁶ Participants will be given the opportunity to review and validate the researchers’ interpretation of findings, a process termed member-checking.²⁶

The quantitative and qualitative data will then be examined and compared for convergence and divergence.^{24 25} The data will be integrated for final interpretations.

Ethics and dissemination

All research staff will be required to complete and maintain College and University requirements for human subject research training, including the Human Research Protection Program and HIPAA. Following consent, participant data will be de-identified by assigning a random numerical code and separated from personally identifying information. Data will be entered, stored and managed in Qualtrics, a HIPAA-compliant secured online database. For those completing paper copies of the surveys, research staff will manually enter the data into Qualtrics. Data will be exported to SPSS statistical software for analysis. The Michigan State University’s Institutional Review Board has reviewed and approved the study (approval date: 5/2/2024; STUDY ID: 00010729). Written informed consent will be obtained from each participant. Study findings will be disseminated through publication in peer-reviewed journals and presentations at academic conferences.

Discussion

Findings from this study will generate a breadth of important new knowledge. The interviews will provide a rich description and in-depth understanding of the various SDOH that contribute to anxiety and depressive symptoms among rural-dwelling FCGs. Further, there will be increased information on the impact of health literacy among rural-dwelling FCGs providing PEOLC. These findings will inform future research and help determine best strategies to emotionally support vulnerable rural-dwelling FCGs.

The expected long-term outcomes are to advance the state of the science related to FCG health and PEOLC in rural communities and to develop nursing interventions to improve emotional health. Further, if health literacy is found to have a mediating effect on the relationship between characteristics/SDOH and anxiety/depressive symptoms, interventions that aim to improve health literacy are warranted.

Given their direct links to patient care and health, expected improvements in FCG health literacy could also translate to better patient outcomes. FCGs with limited health literacy are more likely to delay or forego seeking healthcare for their patient, which affects the ability to receive prompt help.⁴⁸ With improved health literacy, FCGs become more active, engaged and confident in their caregiving, with better emotional health; in turn, patient health outcomes may improve, for example, adequate pain and symptom management.^{15 49} From a systems perspective, limited health literacy may contribute to increased emergency trips to the hospital due to the patient’s unrelieved symptoms, such as pain and/or dyspnoea, potentially avoidable and longer hospitalisations and poor quality of life in the patient’s final days.^{18 50} As a result, there is an increased burden on the healthcare system and expenditures on hospital and healthcare costs.^{51 52}

The proposed study aims to understand factors that may be impacted by SDOH primarily health literacy, which would be a point of intervention to offset unnecessary and/or preventable expenses and negative FCG emotional health outcomes. The data obtained from the study will inform strategies to support the emotional health needs of FCGs and shed light on potential methods for increasing access to care in rural areas.

Limitations

Like most research, the proposed study has limitations. First, the cross-sectional design does not allow us to evaluate changes over time. Additionally, the study incorporates a self-selected convenience sample of rural EOL FCGs. Given reliance on self-report, social desirability may contribute to how participants respond. Further, mixed methods studies require methodological proficiency. Experts with strong backgrounds in statistical analysis and qualitative methods will be consulted to ensure the methodological strength of the study. Lastly, despite rigorous strategies to optimise external validity, the sample may not fully represent the heterogeneity of the rural population.

Implications for future practice and research

Findings from this study can inform practice and research. First, this work draws attention to the need for clinicians to consider SDOH, including health literacy, for FCGs who are often the primary health decision maker. FCGs relieve burden on formal healthcare settings by shifting care to the home environment. The proposed study expands knowledge by examining ways for clinicians to better support rural FCGs as care partners. Clinicians must also be aware of the need to address FCGs’ health and well-being particularly since these FCGs are often older and have their own chronic conditions.

This research expands the landscape of scientific knowledge on rural FCGs, an underexamined group in PEOL caregiving. An examination of FCG personal characteristics and SDOH, with an emphasis on health literacy, offers insight into the end-of-life caregiving experience for those FCGs in rural areas.

Conclusions

Rural FCGs are an important part of our healthcare system’s capacity to provide quality PEOLC as they are first-line liaisons for their patient. Thus, it is vital to recognise and support their significant contribution to home-based PEOLC. Improving access to PEOLC and other essential health services under rural conditions can be supported with consideration of their health literacy needs. Moreover, examining SDOH that impact the provision of home-based PEOLC in rural areas and the emotional health of the FCGs who assist in this care will guide best practices.