LESSONS LEARNED FROM SAVVY EXPRESS: A FEASIBILITY STUDY OF A PSYCHODUCATIONAL INTERVENTION FOR CARE PARTNERS

Abstract An increasing number of families, funders, and community providers seek very brief psychosocial caregiver interventions, yet the evidence for such condensed interventions is not established. Based on the Savvy Caregiver Program, we explored the feasibility, acceptability, and outcome trends for a condensed 3-session version titled, Savvy Express. Based on a single-group, pre- and post-test intervention design, we examined post-intervention and 3-month data on 116 English-speaking racially and ethnically diverse care partners caring family members with Alzheimer’s disease and related dementias. 41% of the sample was non-Latino white and comprised of Latinas, African Americans and Asian American/API. Most care partners were either adult children or spouses caring for someone with AD or other dementia. Over 80% were college educated. Two of three participants completed all 3 classes. Our findings indicate significant improvements in caregiver levels of depressive symptomatology and anxiety, competence, management of the situation, reduction of expectations, making positive comparisons, and reactivity to the family member’s memory behavior. Upwards of 90% would recommend the program to other caregivers. Savvy Express is a brief caregiver intervention with high acceptability and feasibility. Improvements in care partner psychosocial outcomes signal a promising practice to reduce the burden of caregiving. A major focus of the paper focuses on barriers and facilitators to uptake of the study procedures and intervention with community-based partners. Future work is needed to establish the efficacy of Savvy Express across a longer observation period, and with less educated, low-income participants, and limited English-speaking families.

understand the differences between such spouses. Using the Medical Expenditures Panel Survey 2016 Experiences with Cancer Supplement, spousal dyads in which one spouse had cancer were identified (n=670). Cancer survivors reported which family members or friends, if any, provided care during or after their cancer treatment. Survivor and spouse sociodemographic characteristics and spouse psychosocial characteristics including depression (Patient Health Questionnaire-2), distress (Kessler-6), and self-rated health were self-reported in the survey. The proportion of spouses identified as caregivers was calculated and compared with those not identified as caregivers on dyadic characteristics. Multivariable logistic regressions compared spouses' depression, distress, and self-rated health by identification as a caregiver. All analyses employed survey weighting. Most married cancer survivors reported that their spouse was a caregiver (32%) or that they did not have a caregiver (65%); very few did not list their spouse as a caregiver (3%). Survivors who were white (non-Hispanic) or off treatment were less likely to report that their spouse was a caregiver (30% vs. 46%, p0.10; 9.6% vs 11.1%, p>0.10; and 46% vs 50%, p>0.10, respectively). The findings suggest that spouses' experiences may be similar regardless of whether they are considered a caregiver, with implications for research and service delivery. Evidence on the effect of caregiving on cognitive health is mixed. Literature suggests that highly-strained subgroups of CGs experience impairment especially related to depression and negative health behaviors. Conversely, others suggest that sustained higher amounts of caregiving is associated with better cognitive functioning over time. This study compared baseline level and 10-year change in executive function (EF) and episodic memory (EM) in middle aged adults with no caregiving exposure, short (less than one year), average (one to less than five years), and long (five or more years) caregiving exposure. Data came from waves 2 and 3 of the Midlife in the United States study. We statistically matched 265 CGs to 265 non-CGs on baseline sex, age, race, education, marital status, and physical health and controlled for depression and anxiety in analyses. Mixed models estimated fixed-effects of time, depression, and a time by caregiving exposure interaction relative to no caregiving exposure. For EF, CGs with short exposure had higher baseline scores relative to non-CGs (β=.294;p=.023). Though each group declined over time (β=.298;p<.001), caregiving exposure was not related to the rate of decline in EF. Baseline levels of EM were highest in CGs with average exposure (β =.521;p=.011) and long exposure (β=.482;p =.017). However, not all participants declined in EM over 10 years; CGs with long exposure exhibited the only statistically significant decline (β =-.262;p=.026). Results suggest that CG duration is intimately related to EF and EM at baseline, but perhaps only to rate of decline for EM. The sample of this study included older adults aged 60 and above. The respondents were coded as spousal caregivers if they provided activities of daily life (ADL)/instrumental activities of daily life (IADL) assistance to their spouses and did not receive any ADL/IADL help. Ordinary Least Square (OLS) regression analyses were applied. Results: Results from the regression model indicated that spousal caregiving was significantly associated with depressive symptoms (p<0.01). The spousal caregivers showed more depressive symptoms than the non-caregivers. Among the spousal caregivers, the females and those who reported poorer health had more depressive symptoms than their counterparts (p<0.01). Conclusion: The results highlight the negative influence of caregiving on psychological health of spousal caregivers. Caregiver support should be targeted on the older spousal caregivers, especially the wife caregivers. An increasing number of families, funders, and community providers seek very brief psychosocial caregiver interventions, yet the evidence for such condensed interventions is not established. Based on the Savvy Caregiver Program, we explored the feasibility, acceptability, and outcome trends for a condensed 3-session version titled, Savvy Express. Based on a single-group, pre-and post-test intervention design, we examined post-intervention and 3-month data on 116 English-speaking racially and ethnically diverse care partners caring family members with Alzheimer's disease and related dementias. 41% of the sample was non-Latino white and comprised of Latinas, African Americans and Asian American/API. Most care partners were either adult children or spouses caring for someone with AD or other dementia. Over 80% were college educated. Two of three participants completed all 3 classes. Our findings indicate significant improvements in caregiver levels of depressive symptomatology and anxiety, competence, management of the situation, reduction of expectations, making positive comparisons, and reactivity to the family member's memory behavior. Upwards of 90% would recommend the program to other caregivers. Savvy Express is a brief caregiver intervention with high acceptability and feasibility. Improvements in care partner psychosocial outcomes signal a promising practice to reduce the burden of caregiving. A major focus of the paper focuses on barriers and facilitators to uptake of the study procedures and intervention with community-based partners. Future work is needed to establish the efficacy of Savvy Express across a longer observation period, and with less educated, low-income participants, and limited English-speaking families.

TRANSITIONS OF FAMILY CAREGIVERS: A CONCEPTUAL FRAMEWORK FOR END-OF-LIFE CAREGIVING Aimee Fox, 1 and Christine Fruhauf 1 , 1. Colorado State University, Fort Collins, Colorado, United States
Increased longevity and a growing aging population have led to rising numbers of informal, family caregivers. Compared to non-caregivers, caregivers are more stressed and depressed, and report lower levels of self-efficacy, physical health, and subjective well-being. Along the caregiving trajectory, end-of-life (EoL) care is emerging as a distinct phase that is not understood and warrants further research. This presentation will include an integrative literature review on EoL caregivers and propose a model for the transition to caregiving at the end-of-life. The review highlights the differences in strengths and needs of EoL caregivers when compared to non-EoL caregivers and age-related non-caregiving peers. For example, family members providing EoL care are more vulnerable to stress and psychological and physical strain, report low preparedness for EoL care, and often indicate they do not have the knowledge or skills needed for providing care at the EoL. The model for the transition to caregiving at the end-of-life presents new approaches to traditional stress and coping models and a conceptual framework for future EoL caregiving research and intervention efforts. The model highlights the need for expanding the theoretical foundations of EoL caregiving research and guides innovative methods of data collection and thoughtful approaches to the timing, recruitment, and utilization efforts of intervention programs. Ultimately, as researchers identify the needs of EoL caregivers, practitioners can improve caregivers health and well-being not only during this stage, but as they follow their own developmental path beyond loss and bereavement.

SUPPORT AND EDUCATION NEED FULFILLMENT IN INFORMAL DEMENTIA CAREGIVERS
Nicholas James, 1 and Daniel Paulson 1 , 1. University of Central Florida, Orlando, Florida, United States Caregivers vary widely in their need for and utilization of support services, and there are many reasons for this (e.g., time or financial cost, distance, perception of need). This study explored the disparity between services that are desired and those that are utilized, and examined the hypothesis that unmet needs are a determinant of caregiver burden. An online sample of informal caregivers (N=151) responded to a questionnaire containing a list of common caregiver support services: sharing duties, professional transportation, respite care, non-profit community organizations, financial counseling, caregiver education programs, support groups.
Participants were presented with a list of support services and information and asked to indicate A) which they desired and B) which they had received while providing care. Respite care and financial counseling were identified as support that caregivers were unable to obtain, while caregiving education and family/friend support were most commonly fulfilled. A linear regression model controlling for demographic variables was constructed. Unfulfilled support needs accounted for 40.6% of variance in caregiver burden, however the final model included only total ratio of unfulfilled services, transportation services, stress management skills, professional treatment for the caregiver, and behavioral management skills. Results highlight the unique contribution of certain support services in burden reduction. These findings imply a need to improve accessibility to caregiver support, especially those which require payment (e.g., transportation aids and mental healthcare). Further implications and actionable changes related to caregiver support services are discussed.

FACTORS ASSOCIATED WITH CAREGIVER REACTIONS TO PROBLEM BEHAVIORS OF PERSONS WITH ALZHEIMER'S DISEASE
Junrong Shi, 1 and Cathy Scott 1 , 1. University of Tennessee at Chattanooga, Chattanooga, Tennessee, United States Problem behaviors among persons with Alzheimer's Disease (AD) can have a major impact on caregivers. However, caregiver's subjective reactions to the problem behaviors have a stronger impact on caregivers than the objective frequency of problem behaviors (Robinson et al., 2001). This study aims to examine the factors associated with caregiver's subjective reactions to problem behaviors. Data were collected from a sample recruited from community agencies served AD caregivers in the southeastern region of the country (N=109). The caregivers' reactions to problem behaviors were measured by the Revised Memory and Behavior Problem Checklist (Teri et al., 1992). Three subscale scores were used to measure the reactions to the behaviors related to memory loss, depression and disruption. Multivariate regression models were conducted including gender, race, employment, living arrangement, knowledge about the disease, resilience for caregivers; and ADL and IADL functioning, and frequency of problem behaviors for care recipients. Race (B=-.162; p<.05) and frequency of care-recipient problem behaviors (B= 0.733; p<.001) were significantly associated with caregiver's reaction to problem behaviors. Caregiver's knowledge about the disease (p<.01) only influence their reactions to memory loss problems but not for disruption and depression problems. African American caregivers had fewer reactions to disruption (p<.01) and more reactions to depression problems (p=0.06) than white caregivers, but no difference between the two groups in their reactions to memory loss problems. Personal resilience was not associated with reactions to any problem behaviors. Intervention should be tailored to the needs of caregivers to deal with behaviors of a person with AD.