The development of knowledgeable nursing assistants as creative caregivers (KNACC).

Certified nursing assistants (CNAs) provide 80% of direct care in long-term care settings and are critical to maintaining resident well-being. Arts-based approaches to enhancing meaningful engagement have the potential to empower CNA ownership in the process of improving patient-centered care. We held a series of focus groups with CNAs (n = 14) to adapt arts-based creative caregiving (CCG) techniques for use in long-term care. Iterative revisions focused on CCG techniques, factors influencing implementation, and usability. The Knowledgeable Nursing Assistants as Creative Caregivers (KNACC) manual developed from the adapted CCG describes training guidelines and instructions to facilitate CNA use of creative caregiving techniques in direct care.

Pilot testing the knowledgeable nursing assistants as creative caregivers (KNACC) guide.

Knowledgeable Nursing Assistants as Creative Caregivers (KNACC) was developed to train certified nursing assistants (CNAs) to apply arts-based techniques to enhance care and support to older adults in long-term care (LTC) settings. We piloted (n=8) KNACC techniques for use in LTC and assessed its potential for influence on the CNA outcomes of structural empowerment and job satisfaction. During preliminary implementation, CNAs working in memory care units were more open to applying all techniques. The techniques in KNACC have the potential to improve CNA skills in working with long-term care residents.

Providing home hospice care for LGBTQ+ patients and caregivers: Perceptions and opinions of hospice interdisciplinary care team providers.

OBJECTIVES: Hospice patients and caregivers who are members of sexual and gender minority groups (i.e., LGBTQ+) have reported experiencing unmet needs at end of life (EOL). Negative experiences often stem from challenging interactions with healthcare providers due to ineffective or poor communication and providers' heteronormative assumptions and biases. Few studies, however, examine hospice care team (HCT) providers' knowledge, experience, and opinions related to EOL care for LGBTQ+ patients and caregivers despite this being identified as a gap in competency and education. We sought to examine HCT providers' perceptions regarding (1) awareness of LGBTQ+ patients and caregivers; (2) knowledge of specific or unique needs; and (3) opinions on best care and communication practices. METHODS: Six focus groups conducted with HCT providers (n = 48) currently delivering hospice care in three US states were audio-recorded and transcribed. Data were content coded (κ = 0.77), aggregated by topical categories, and descriptively summarized. RESULTS: Participants were mostly white and non-Hispanic (n = 43, 89.6%), cisgender female (n = 42, 87.5%), heterosexual (n = 35, 72.9%), and religious (n = 33, 68.8%); they averaged 49 years of age (range 26-72, SD = 11.66). Awareness of LGBTQ+ patients and caregivers depended on patient or caregiver self-disclosure and contextual cues; orientation and gender identity data were not routinely collected. Many viewed being LGBTQ+ as private, irrelevant to care, and not a basis for people having specific or unique EOL needs because they saw EOL processes as universal, and believed that they treat everyone equally. Providers were more comfortable with patients of lesbian or gay orientation and reported less comfort and limited experience caring for transgender and gender-diverse patients or caregivers. SIGNIFICANCE OF RESULTS: Many HCT members were unaware of specific issues impacting the EOL experiences of LGBTQ+ patients and caregivers, or how these experiences may inform important care and communication needs at EOL.

"Being careful": How much caution is enough? The relational context of fear of older adult falling in older adult-family caregiver dyads.

Family caregivers play an important role in coping with older adult falls; however, their perspectives on fear of older adult falling are lacking from the falls prevention literature. A mixed-method design (N=25 dyads) with interview and survey data examined linguistic characteristics and coping strategies used by older adult and family caregiver dyads to manage fear of older adult falling. Fear of older adult falling consisted of both affective (e.g., worry) and cognitive (e.g., cautious) properties. Family caregivers more frequently used affective words and first-person plural pronouns ("we" language) when talking about fear of older adult falling, while older adults more frequently used cognitive and first-and-second person singular pronouns ("I", "you"). The concept of "being careful" was shared within dyads. However, dyad partners differed in their perspectives of what constituted "being careful" and the possibilities of future falling. Findings suggest that the need for family-centered interventions to prevent falls are needed.

Engaging Sexual and Gender Minority Older Adults to Elicit Concerns and Recommendations for Communicating Care Preferences in Long-Term Care: Focus Group Findings.

OBJECTIVES: 1) Describe sexual and gender minority (SGM) older adults' current practices and concerns for communicating end-of-life (EOL) and daily care preferences in long-term care (LTC); 2) Elicit ideas about adapting a video-based intervention to facilitate communication. METHODS: After consulting a community advisory board, we conducted two focus groups with SGM older adults ≥55 years of age recruited from a community-based service organization (n = 4) and a continuing care retirement community (n = 9). We audio-recorded, transcribed, and coded data using directed content analysis methods and summarized results descriptively. RESULTS: Most participants were cisgender (n = 12), female (n = 11), lesbian (n = 10), White and non-Hispanic (n = 13); mean age was 70.62. Participants' concerns included discrimination, autonomy, chosen family, and community; they linked daily care preferences with personhood and quality of life. They advocated for building intentional community-based support networks to help peers discuss preferences and produce videos before LTC transition and ensure compliance after. CONCLUSIONS: For SGM older adults, asserting and protecting their full personhood, through daily care preferences, is essential to quality of life in LTC. CLINICAL IMPLICATIONS: Video-recorded statements of daily preferences, facilitated and supported by a peer network, could promote culturally competent, person- and community-centered care for SGM older adults in LTC settings.

What contextual factors account for anxiety and depressed mood in hospice family caregivers?

OBJECTIVE: End-of-life caregiving is associated with poorer mental health compared with other caregiving. The objective of this study was to examine the association between contextual characteristics and appraisal factors on family caregivers' mental health and well-being. METHODS: Family hospice caregivers were recruited across four states using a non-probabilistic sampling approach. This study analyzed contextual (demographic, caregiving, economic) and appraisal factors (Medical Outcomes Study Social Support Survey, Zarit Burden Interview) on caregivers' anxiety and depression (Hospital Anxiety and Depression scale, and positive affect and well-being (Positive Affect and Well-being Scale). Hierarchical linear regression models were generated in SPSS version 24. RESULTS: Data from 102 family caregivers were analyzed. On average, participants were 58.93 years of age (SD = 14.24), mostly female (72.55%), spouses/partners (51.96%), and non-Hispanic White (78.43%). Most (75.49%) described their financial situation as comfortable or more than adequate. Younger age (B = -0.11, 95% CI = -0.18 to -0.05) and increased caregiving burden (B = 0.18, 95% CI = 0.09 to 0.27) were associated with increased anxiety, while lower perceived financial adequacy (B = -1.19, 95% CI = -2.07 to -0.32), lower social support (B = -0.04, 95% CI = -0.06 to -0.01), and increased caregiving burden (B = 0.15, 95% CI = 0.08-0.22) were associated with worsened depression. Greater social support (B = 0.10, 95% CI = 0.05-0.14) and lower caregiving burden (B = -0.19, 95% CI = -0.32 to -0.07) were associated with greater positive affect and well-being. CONCLUSIONS: Findings suggest significant impact of contextual factors on mental health and well-being, and support the need for holistic assessment of hospice caregivers' wellbeing and programs and policies providing social services and economic support to caregivers.

Posttraumatic growth and life course stress predict dyadic wellbeing among sexual and gender minority (SGM) and non-SGM couples facing cancer.

OBJECTIVE: This study examined: (1) Differences among sexual and gender minority (SGM) and non-SGM couples' life course stress, posttraumatic growth (PTG), individual, and dyadic wellbeing while facing cancer, (2) The predictive ability of PTG and life course stress on wellbeing for each dyad member, and (3) The predictive ability of dyad-level PTG and dyad-level life course stress on dyadic wellbeing. METHODS: Thirty SGM and 30 non-SGM dyads (N = 60) completed measures assessing demographics, life course stress, PTG, individual, and dyadic wellbeing. Regression and multilevel models tested predictive hypotheses. RESULTS: Participants were 56.3 years old on average (SD = 13.6) and were together for 24.2 (SD = 14.9) years. SGM participants reported greater life course stress and higher scores on the Depression, Anxiety, and Stress Scales (DASS-21) than non-SGM participants. A dyad members' higher PTG predicted their partners' higher DASS-21 score, dyad members' higher life course stress predicted their own higher DASS-21 score, and patients' higher life course stress predicted their partners' higher DASS-21 score. Greater dyadic PTG predicted greater dyadic wellbeing. CONCLUSIONS: SGM and non-SGM couples experience PTG equally despite SGM couples' greater life course stress and higher DASS-21 scores. Future research is needed to explore how PTG may affect individuals and couples differently.

Cancer content and social media platform influence young adult cancer caregivers' social support on social media.

PURPOSE: To determine how social media platform and cancer content is associated with the presence of social support in responses to young adult cancer caregivers' (YACC) posts. DESIGN: We retrospectively collected YACC's Facebook and/or Instagram posts and all responses from the first six months of caregiving. SAMPLE: Eligible YACC were 18-39, caring for a cancer patient diagnosed 6 months-5 years prior, spoke English, and used social media weekly. METHODS: Social media posts and responses were manually coded for five social support types, then transformed to depict the proportion of responses per post representing each type of support. Using mixed-effects models, we compared the distributions of responses with social support types by platform (Facebook vs. Instagram) and cancer content (no vs. yes). FINDINGS: More responses contained emotional support on Instagram than Facebook (B = 0.25, Standard Error (SE)=0.09, p = 0.007). More responses with cancer content contained -validation support (B = 0.20, SE = 0.07, p = 0.002), but fewer contained emotional (B=-0.17, SE = 0.07, p = 0.02) and instrumental support (B=-0.06, SE = 0.02, p = 0.001) than posts without cancer content. CONCLUSIONS: Studying the responsiveness of social media followers by platform and cancer content provides a foundation for intervention development. IMPLICATIONS FOR PSYCHOSOCIAL PROVIDERS: Emphasizing the suitability of different social media platforms for particular support seeking behaviors is essential.

Young adult cancer caregivers' exposure to cancer misinformation on social media.

BACKGROUND: The objective of this study was to describe young adult cancer caregivers' exposure to cancer misinformation on social media. METHODS: Eligible participants were 18 to 39 years old, used social media weekly, and cared for an adult patient with cancer diagnosed 6 months to 5 years before (N = 21). Recruitment occurred from August 2017 to June 2018 in person at oncology clinics and online. Semistructured telephone interviews were recorded and analyzed with grounded theory methods through 2 rounds of coding (κ = 0.88). RESULTS: Caregivers were on average 29.1 years old, female (76.2%), non-Hispanic White (90.5%), college educated (57.1%), and caring for a spouse/partner (47.6%). Caregivers were exposed to a variety of cancer misinformation topics; some felt uncertain in their caregiving ability when confronted with cancer misinformation. Caregivers crosschecked online sources and consulted trusted individuals (eg, family, friends, and health care providers) to determine the quality of cancer-related information and manage cancer misinformation. CONCLUSIONS: Cancer misinformation on social media may influence caregiving behaviors and decision making. Whether caregivers correctly and consistently identify cancer misinformation on social media is unknown. Supporting digital literacy to mitigate cancer misinformation on social media may improve young cancer caregivers' access to credible cancer information. LAY SUMMARY: Cancer misinformation on social media affects young adult cancer caregivers' perceptions of their social support network. Caregivers differ in how they evaluate cancer information for quality and in how this information influences their health behaviors and caregiving decisions.

Palliative and End-of-Life Care for Sexual and Gender Minority Cancer Survivors: a Review of Current Research and Recommendations.

PURPOSE OF REVIEW: Sexual and gender minority (SGM) cancer survivors (post-diagnosis to end-of-life) can face inadequate services along the cancer care continuum, including palliative and end-of-life (EOL) care. A growing literature base calls for more research on factors influencing palliative and EOL care access, outcomes, and distinct needs of SGM cancer survivors and caregivers. We analyzed peer-reviewed articles published 2015-present to identify trends. RECENT FINDINGS: Ten articles were SGM-focused or inclusive, cancer-focused, and included substantive discussion of palliative and/or EOL care. Four were research studies (three case studies and one qualitative interview study) and six were literature reviews. Recurrent topics included disparities in cancer risk, access, and care; essential cultural and clinical competencies; and need for professional and organizational standards and policies addressing interpersonal and institutional discrimination and inclusion. Provision of equitable, competent palliative and EOL care depends on continued advancements in research, translated into person-centered approaches to care. We discuss implications of findings for improving palliative and EOL care for SGM cancer survivors.

The Me in We dyadic communication intervention is feasible and acceptable among advanced cancer patients and their family caregivers.

BACKGROUND: Advanced cancer affects the emotional and physical well-being of both patients and family caregivers in profound ways and is experienced both dyadically and individually. Dyadic interventions address the concerns of both members of the dyad. A critical gap exists in advanced cancer research, which is a failure of goals research and dyadic research to fully account for the reciprocal and synergistic effects of patients' and caregivers' individual perspectives, and those they share. AIM: We describe the feasibility and acceptability of the Me in We dyadic intervention, which is aimed at facilitating communication and goals-sharing among caregiver and patient dyads while integrating family context and individual/shared perspectives. DESIGN: Pilot study of a participant-generated goals communication intervention, guided by multiple goals theory, with 13 patient-caregiver dyads over two sessions. SETTING/PARTICIPANTS: Patients with advanced cancer and their self-identified family caregivers were recruited from an academic cancer center. Dyads did not have to live together, but both had to consent to participate and all participants had to speak and read English and be at least 18 years or age. RESULTS: Of those approached, 54.8% dyads agreed to participate and completed both sessions. Participants generated and openly discussed their personal and shared goals and experienced positive emotions during the sessions. CONCLUSIONS: This intervention showed feasibility and acceptability using participant-generated goals as personalized points of communication for advanced cancer dyads. This model shows promise as a communication intervention for dyads in discussing and working towards individual and shared goals when facing life-limiting or end-of-life cancer.

Patterns of stress and support in social support networks of in-home hospice cancer family caregivers.

Family caregivers of home hospice cancer patients often experience burden and distress, which can be mitigated by perceived social support. However, less attention has been paid to the non-family sources of support within social networks, or to how sources of support may also be sources of stress. We describe support and stress in social networks of hospice family caregivers and identify caregiving characteristics associated with classes identified in our data. We collected demographic and psychosocial self-report data from family caregivers providing in-home hospice care for advanced cancer patients (N = 90). Caregivers also reported perceived support and stress from specific family and non-family relationships. We identified three classes with unique patterns of stress and support within caregivers' support networks using a latent class analysis. Classes include: 1) high support, low stress across family and non-family network members ("supportive"; 53% of caregivers); 2) high support, high stress across family and non-family network ("ambivalent maximizers"; 26%); and 3) high support, high stress across family network only ("family-focused ambivalent"; 21%). Caregivers in the ambivalent maximizer class reported more burden than caregivers in the supportive class (p = .024). This is one of the first studies to systematically explore the role of non-family support, as well as how stress and support co-occur within relationships and across networks. As informal support networks of hospice family caregivers are complex and multifaceted, understanding the patterns of support and stress across various network members is essential to offer services to more effectively manage caregiver burden.

Young adult cancer caregivers' use of social media for social support.

OBJECTIVE: To describe how young adult cancer caregivers (YACC) use social media for social support during a cancer experience. METHODS: Eligible YACC were 18 to 39 years, used Facebook and/or Instagram at least once per week, and cared for an adult cancer patient diagnosed 6 months to 5 years prior (N = 34). Recruitment of a cross-sectional sample occurred through oncology clinics in Utah and online advertising by caregiving and cancer organizations from September 2017 to June 2018. Semi-structured telephone interviews were recorded, transcribed, iteratively coded, and qualitatively analyzed, yielding four categories concerning how YACC use social media. RESULTS: Caregivers were most commonly spouses aged 29 years on average (range 21-38); cancer patients were 37 years (range 19-76). Analysis yielded four distinct yet related categories: Category 1: Posting about cancer on social media often begins as a strategy for YACC to efficiently provide updates about the cancer patient. Category 2: Caregivers who actively post on social media experience a variety of different functional social supports to which they otherwise would not have access. Category 3: Posting about cancer online presents an opportunity for negative consequences. Category 4: Potential for negative consequences influences how some caregivers use social media. CONCLUSIONS: Supportive services, including social media-based supports, are needed for YACC in formats that are convenient for them as they balance their caretaking duties with their daily lives.

Fear of older adult falling questionnaire for caregivers (FOAFQ-CG): Evidence from content validity and item-response theory graded-response modelling.

AIMS: (1) Determine the content validity of the Fear of Older Adult Falling Questionnaire-Caregivers using a panel of gerontological experts and a target sample of family caregivers (Stage 1) and (2) Examine the response patterns of the Fear of Older Adult Falling Questionnaire-Caregivers and compare it with older adult version of Fear of Falling Questionnaire Revised using graded-response modelling (Stage 2). DESIGN: Cross-sectional mixed-method design. METHODS: Five content experts and 10 family caregivers were involved in the Stage 1 study and 53 family caregiver-older adult dyads (N = 106) were included in the Stage 2 study. The content-validity index and graded-response modelling were used to analyse data. RESULTS: Among experts, the Fear of Older Adult Falling Questionnaire-Caregivers content-validity index for relevancy, importance, and clarity of individual items and total scale ranged from 0.60-1.00 and from 0.77-0.87, respectively. Among family caregivers, the ratings of the item and scale level content-validity index for relevancy, importance, and clarity ranged from 0.90-1.00 and from 0.95-0.97, respectively. Combining feedback from both groups, we revised one item. Subsequently, the graded-response modelling revealed that a 1-factor, 3-item version of the Fear of Older Adult Falling Questionnaire-Caregivers had acceptable psychometric properties. CONCLUSIONS: The brief 3-item version of the Fear of Older Adult Falling Questionnaire-Caregivers is promising for assessing caregivers' fear of their older adult care recipient falling. IMPACT: A significant concern for family caregivers is fearing that older adult care recipients will fall, but a lack of validated measures limits the study of this phenomena. A 3-item version of the Fear of Older Adult Falling Questionnaire-Caregivers has the potential to identify family caregivers with high fear of older adult falling so that fall risk can be appropriately assessed and addressed.

Mining twitter to explore the emergence of COVID-19 symptoms.

BACKGROUND: The Centers for Disease Control and Prevention (CDC) in United States initially alerted the public to three COVID-19 signs and symptoms-fever, dry cough, and shortness of breath. Concurrent social media posts reflected a wider range of symptoms of COVID-19 besides these three symptoms. Because social media data have a potential application in the early identification novel virus symptoms, this study aimed to explore what symptoms mentioned in COVID-19-related social media posts during the early stages of the pandemic. METHODS: We collected COVID-19-related Twitter tweets posted in English language between March 30, 2020 and April 19, 2020 using search terms of COVID-19 synonyms and three common COVID-19 symptoms suggested by the CDC in March. Only unique tweets were extracted for analysis of symptom terms. RESULTS: A total of 36 symptoms were extracted from 30,732 unique tweets. All the symptoms suggested by the CDC for COVID-19 screening in March, April, and May were mentioned in tweets posted during the early stages of the pandemic. DISCUSSION: The findings of this study revealed that many COVID-19-related symptoms mentioned in Twitter tweets earlier than the announcement by the CDC. Monitoring social media data is a promising approach to public health surveillance.

Supporting home hospice family caregivers: Insights from different perspectives.

ABSTRACTObjective:Our intention was to describe and compare the perspectives of national hospice thought leaders, hospice nurses, and former family caregivers on factors that promote or threaten family caregiver perceptions of support. METHOD: Nationally recognized hospice thought leaders (n = 11), hospice nurses (n = 13), and former family caregivers (n = 14) participated. Interviews and focus groups were audiotaped and transcribed. Data were coded inductively, and codes were hierarchically grouped by topic. Emergent categories were summarized descriptively and compared across groups. RESULTS: Four categories linked responses from the three participant groups (95%, 366/384 codes): (1) essentials of skilled communication (30.6%), (2) importance of building authentic relationships (28%), (3) value of expert teaching (22.4%), and (4) critical role of teamwork (18.3%). The thought leaders emphasized communication (44.6%), caregivers stressed expert teaching (51%), and nurses highlighted teamwork (35.8%). Nurses discussed teamwork significantly more than caregivers (z = 2.2786), thought leaders discussed communication more than caregivers (z = 2.8551), and caregivers discussed expert teaching more than thought leaders (z = 2.1693) and nurses (z = 2.4718; all values of p < 0.05). SIGNIFICANCE OF RESULTS: Our findings suggest differences in priorities for caregiver support across family caregivers, hospice nurses, and thought leaders. Hospice teams may benefit from further education and training to help cross the schism of family-centered hospice care as a clinical ideal to one where hospice team members can fully support and empower family caregivers as a hospice team member.

Social support needs: discordance between home hospice nurses and former family caregivers.

OBJECTIVE: The two goals of our study were to (1) identify which of five types of social support (Informational, Esteem, Emotional, Tangible, Belonging) are most cited by hospice nurses and family caregivers and (2) determine the match in perception of support needs. METHOD: As part of a larger multiphase project, focus groups were conducted with former family caregivers and hospice nurses to discuss their experiences of home hospice care and to gather their opinions on the important issues involved in that care. Transcripts of focus group discussions were coded for support type (Informational, Esteem, Emotional, Tangible, Belonging) based on definitions from the literature. Nurse and caregiver data were compared to assess for potential match. RESULTS: Analysis of coded data suggested that nurses see different types of support to be needed in equal measure across their caseloads, while caregivers expressed priorities for some types of support. Illustrative examples of each type of support are provided and discussed. SIGNIFICANCE OF RESULTS: Because matching support provided with the type of support desired has been linked to improved physical and psychological outcomes, it is important to focus on this match in healthcare populations particularly vulnerable to psychological stress, including family caregivers of home hospice patients. This research has implications for interventions to match support provision to caregiver needs, or for education for home hospice providers to ensure that they are not only sensitive to the possibility of a broad range of needs but also to the necessity to tailor care to those needs.

Blog text about female incontinence: presentation of self, disclosure, and social risk assessment.

BACKGROUND: Female urinary incontinence is a significant health concern that often remains undisclosed to healthcare providers, often because of embarrassment and fear of a negative response. OBJECTIVES: The purpose of this study was to explore communication patterns found in blog text regarding self-presentation by women living with incontinence for clues to facilitate disclosure. METHODS: This is a descriptive, empirical study of publically archived blog text (years inclusive 2006-2011; N = 16,629 words from 19 bloggers) by women communicating about urinary incontinence, utilizing methods of content analysis. RESULTS: Valence and word concordance analysis showed that words indicative of facilitating social connections were used more often in proximity with close contacts (Z = -2.68, p = .004) and words indicative of blocking social connections were used more often in proximity with community contacts (Z = -2.97, p = .002). Differences between descriptions of facilitative and blocking reactions from healthcare providers were not significant (Z = -0.28, p = .39). Cluster analysis indicated a decreasing level of negatively charged descriptors of incontinence-related communications as the context moved from the hidden self to close contacts and then to the public sphere. Word frequency analysis identified a pattern in the blog text about urinary incontinence of self-presenting as otherwise fit, healthy, and competent. DISCUSSION: Study results suggest that any report of incontinence concerns, including joking or casual references, should be addressed because women may not disclose the degree to which symptoms affect their psychosocial health. Further research is needed to explore whether providers might facilitate disclosure of urinary incontinence by first acknowledging the woman's strengths, thereby creating a sense of safety and acceptance.

The grief experience of prison inmate hospice volunteer caregivers.

Correctional institutions are obligated to provide end-of-life care to a population with complex medical needs. Prison hospices are increasingly being formed to address this demand. Few empirical studies have examined the impact of caring for dying inmates on the hospice inmate volunteers, who, in several prison health care systems, provide direct care. In this study, experiences of the inmate hospice volunteers with death were investigated to illuminate their grief processes. Understanding the bereavement needs of hospice volunteers and how prison hospice volunteers navigate grief and remain committed to providing excellent hospice care can inform the grief processes and practices of hospice care professionals.

Longitudinal Analysis of Cancer Family Caregiver Perception of Sleep Difficulty During Home Hospice.

BACKGROUND: Sleep difficulty in caregivers is associated with poor physical and psychological outcomes. The purpose of this study was to describe family caregivers' perception of sleep difficulty through the hospice trajectory after a cancer diagnosis as predicted by age, sex, self-report of anxiety or depression, and cohabitation. METHODS: We conducted a secondary analysis of longitudinal data using multilevel modeling with nested model comparisons. Beginning with an unconditional growth model, predictors were added to nested models to test differential impact. RESULTS: Caregivers (n = 164) were predominately white (n = 160; 97%) and female (n = 113, 69%). We hypothesized that age, sex, history of anxiety or depression, and cohabitation would predict sleep difficulty. The cohabitation predictor model was a statistically significant model for caregiver perception of sleep difficulty that worsened throughout hospice caregiving (b = .184, χ2 = 7.199, P = 0.027) but age, sex, and history of depression or anxiety did not improve model fit. CONCLUSION: Our findings indicate that family caregivers who cohabitate exhibit increased perception of sleep difficulty over the course of hospice. Future studies and interventions for hospice family caregivers' sleep should consider cohabitation between the patient and the caregiver as a significant predictor of sleep difficulty to observe and potentially mediate the negative outcomes associated with caregiver sleep difficulty. Further, determining the underlying reasons for sleep difficulty in cohabitation (e.g., patient symptoms or treatments) should be explored.