Community REACH: An Implementation of an Evidence- Based Caregiver Program

Transcription

1 The Gerontologist cite as: Gerontologist, 2018, Vol. 58, No. 2, e130 e137 doi: /geront/gny001 Intervention Research Community REACH: An Implementation of an Evidence- Based Caregiver Program Sara J. Czaja, PhD, 1, * Chin Chin Lee, MSPH, 2 Dolores Perdomo, PhD, 1 David Loewenstein, PhD, 1 Marina Bravo, LCSW, 3 Jerad H. Moxley, PhD, 2 and Richard Schulz, PhD 4 1 Department of Psychiatry and Behavioral Sciences/Center on Aging and 2 Center on Aging, University of Miami Miller School of Medicine, Florida. 3 Neurobehavioral Associates, Doral, Florida. 4 University Center for Social and Urban Research, University of Pittsburgh, Pennsylvania. *Address correspondence to: Sara J. Czaja, PhD, Department of Psychiatry and Behavioral Sciences/Center on Aging, University of Miami Miller School of Medicine, 1695 N.W. 9th Avenue, Suite 3208 (D-101), Miami, Florida sczaja@med.miami.edu Received: September 28, 2017; Editorial Decision Date: January 9, 2018 Decision Editor: Suzanne Meeks, PhD Abstract Background and Objectives: Family caregivers (CGs) are critical to the provision of long-term services and support for older adults. Numerous intervention programs to alleviate CG distress have been developed and evaluated yet few have been implemented in community settings. This paper describes and presents outcomes from Community REACH, a community implementation of the evidence-based Resources for Enhancing Alzheimer s Caregiver Health (REACH) II program. Research Design and Methods: Community REACH involved a partnership between REACH II investigators and United HomeCare Services (UHCS), a nonprofit home health organization that provides home health, personal care, companion, and respite services. The intervention program, an adapted version of an evidence-based program, was a 6-month multicomponent psychosocial intervention, which involved six individual face-to-face and six individual telephone sessions, and telephone support groups. One hundred and forty-six CGs who were primarily female (76%) and Latino, and providing care for an individual with Alzheimer s disease (AD) were enrolled. Program effectiveness was assessed by examining changes in perceived social support, burden, and depression, and CG self-efficacy. Results: At 6 months, CGs reported significant decreases in depression, burden, being and bothered by the care recipient s memory problems. There was also a significant decline in the number of CGs at risk for clinical depression. These improvements were maintained at 12 months and there was an increase in feelings of social support. Discussion and Implications: The findings indicate that evidence-based CG programs can be successfully implemented in community settings and benefit CGs of AD patients. A continued partnership between the program developers and community partners is key to implementation success. Keywords: Alzheimer s disease, Caregiving, Intervention Family caregivers (CGs) represent a critical source of support for our increasing population of older adults. Currently, about 17.7 million adults in the United States are providing care to an older adult who has a chronic disease or disability (The National Academies of Sciences, Engineering, Medicine, 2016). Many CGs are providing care for an older adult who has Alzheimer s disease (AD) or dementia. An estimated 5.4 million Americans have AD, 96% of whom are aged 65+ and cared for by family members (Alzheimer s Association, 2017). This is concerning as The Author(s) Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please journals.permissions@oup.com. e130

2 The Gerontologist, 2018, Vol. 58, No. 2 e131 the prevalence of persons with AD is projected to increase if no cure is found given the aging of the baby boomers and increase in the oldest old cohort (85+). Caregivers perform a variety of tasks including care coordination; assistance with bathing, dressing, and finances; and medical/nursing tasks such as medication and symptom management. Unfortunately, most CGs are unprepared for the caregiving role and although CGs can reap positive benefits from the caregiving experience, many experience adverse physical, social and emotional consequences (The National Academies of Sciences, Engineering, Medicine, 2016). Caregivers are at higher risk than non- CGs for depression, burden, stress (Schulz, O Brien, Bookwala, & Fleissner, 1995), adverse health outcomes (Capistrant, 2016), and financial distress (The National Academies of Sciences, Engineering, Medicine, 2016) especially when caregiving demands are high and extends for long time periods. Numerous studies have been conducted to develop and evaluate interventions for family CGs. Reviews of this literature (Gitlin & Hodgson, in press; The National Academies of Sciences, Engineering, Medicine, 2016) indicate that on average these studies have resulted in significant small to moderate improvements in CG outcomes. Despite the recent push to promote community-based adoption of evidence-based programs, few of these programs have been implemented in practice and adoption is painstakingly slow (Gitlin & Czaja, 2015; Gitlin, Marx, Standley, & Hodgson, 2015). Further, there is little evaluation of these programs when implemented, especially with minority populations. One of the few exceptions is the REACH VA study (Nichols, Martindale-Adams, Burns, Graney, & Zuber, 2011), which represented the first clinical translational of the REACH II program (Belle et al., 2006), a proven behavioral intervention program for CGs of persons with dementia, across multiple states and Department of Veteran Administration (VA) facilities. REACH VA has been adapted to facilitate implementation. For example, in current implementations the number of intervention sessions has been reduced. The results indicate that CGs who receive the intervention experienced significant decreases in burden and anxiety (Nichols et al., 2011, 2016). The results also demonstrate that an evidence-based intervention can be successfully implemented in clinical practice settings by staff affiliated with the facilities. However, in REACH VA, the implementation was limited to the VA population within a closed care setting and not the community at large. REACH II was also modified for use in four Area Agencies on Aging in Alabama (Burgio et al., 2009). The findings indicated that the intervention (REACH OUT) had a positive impact of CG outcomes such as burden, depression, and social support. Similarly, Gitlin and colleagues (Gitlin, Reever, Dennis, Mathieu, & Hauck, 2006) demonstrated that an intervention, Adult Day Services Plus (ADS Plus) could be successfully implemented within adult day care centers. The intervention was a low-cost care management intervention and designed to enhance family CG well-being, increase service utilization, and decrease nursing home placement of impaired older adults. Teri and colleagues (Teri, McCurry, Logsdon, & Gibbons, 2005) demonstrated that community-based consultants could successfully implement a behavior intervention, STAR_C, for CGs of persons with dementia and that CGs who received the intervention showed improvements in depression, burden, and quality of life. In this study, we examined if the REACH II intervention could be implemented within a community organization, United Homecare Services, (UHCS), a nonprofit home health and community care organization serving predominately low-income, Hispanic older adults in the South Florida community. The intervention was designed to impart family CGs of adults with AD with knowledge about the disease and caregiving, support, and skills to improve their ability to provide care. The intervention was theory driven and based on a variant of the stress process model (Perlin, Mullan, Semple, & Skaff, 1990; Schulz, Gallagher- Thompson, Haley, & Czaja, 2000), which recognized the importance of contextual (i.e., the physical environment) and objective factors such as the behavioral symptoms of the patient in defining the caregiving experience and that multiple pathways can contribute to CG outcomes such as burden. The articulation of this model provided a conceptual framework to guide the intervention, which was multicomponent and targeted five areas along the pathway of CG risk: problem behaviors, social support, education and skill building, depression and self-care, and preventive health practices. The intervention activities were directed at each of these risk areas. For example, activities targeting CG depression included the provision of information and mood management techniques. Current models and findings of implementation science (i.e., Feldstein & Glasgow, 2008; Czaja, Valente, Nair, Villamar, & Brown, 2016), which stress the importance of considering the contextual characteristics of the implementation setting and buy in from the stakeholders guided implementation of the intervention. Thus, the study involved a partnership between REACH II investigators and UHCS. The investigators worked with UHSC personnel to refine the intervention, provided training and clinical oversight for UHCS personnel, and assumed responsibility for program evaluation. The sample was primarily minority CGs. Because Community REACH was designed as a community translation of an evidence-based program (Belle et al., 2006), there was no control group. Program effectiveness was assessed using a pre- post-test design that included a baseline assessment and an assessment immediately following treatment (6 months post randomization). Unlike REACH II, we also examined maintenance of treatment effects with a follow-up assessment 6 months post-treatment (12 months post randomization). We predicted that Community REACH would improve CG outcomes on burden and depression.

3 e132 The Gerontologist, 2018, Vol. 58, No. 2 Method Setting UHCS enrolled and was the direct service provider for the intervention and conducted the assessments. UHCS staff were trained in the delivery of the intervention by the REACH II investigators and provided with a detailed intervention manual. UHCS staff included a clinical director, project coordinator, and three interventionists/assessors. Follow-up assessments were done by interventionists/assessors who were not involved in intervention delivery. UHSC staff had at least a bachelor s degree and experience with patients with AD and family CGs. The clinical director led weekly clinical meetings of the UHCS team. The REACH II investigators were available for consultations. The project also included a Steering Committee that met semiannually and included: the REACH II investigators, representatives from UHSC and agencies serving older adults and CGs (e.g., Department of Elder Affairs, Alzheimer s Association); and local memory disorder clinics. Participants UHSC staff recruited and screened the participants. Caregiver inclusion criteria were: (a) being 21 years of age or older; (b) living with or share cooking facilities with the care recipient; (c) providing care for a relative with diagnosed ADRD for at least 4 hr per day for at least the past 6 months; (d) reporting distress associated with caregiving at the baseline assessment (i.e., feeling overwhelmed, angry or frustrated, isolated from family/friends, a frequent need to cry, moderate to high levels of general stress, or that their health had declined); (e) not involved in another CG intervention study; and (f) not having an illness preventing their participation for the study duration. Outcome Measures The efficacy of the intervention was evaluated by examining changes between baseline and 6 months in measures of CG depression (10-Item Center for Epidemiological Depression Scale [CESD-R-10] (Andresen, Malmgren, Carter, & Patrick, 1994)), Overall Burden (Burden Interview [Bedard et al., 2001; Zarit, Orr, & Zarit, 1985]), and obervable behavioral problems in the CR and the CG reaction to the behavior relative to how upsetting the behavior was to the CG. The measure includes three subscales: memoryrelated problems, affective distress, and disruptive behaviors (Revised Memory and Behavior Problems Checklist [Teri et al., 1992]), Social Support (Barrera, Sandler, & Ramsay, 1981; Krause, 1995; Krause & Markides, 1990), and Caregiving Self- Efficacy (Steffen, McKibbin, Zeiss, Gallagher-Thompson, & Bandura, 2002). The Caregiver Self-Efficacy Scale measures three domains of caregiving self-efficacy: Obtaining Respite, Responding to Disruptive Patient Behaviors, and Controlling Upsetting Thoughts. The Social Support Questionnaire included three subscales reflecting received support, satisfaction with support and negative (nonsupportive) interactions. For this analysis, similar to REACH II (Belle et al., 2006), we combined the three subscales to derive an overall measure of social support. Given that we used the 10-item CESD-R-10, we used a cutoff score of 15 or higher (equivalent to 30 on the full 20-item scale, which reflects extremely high levels of reported depressive symptoms) for designating individuals at risk for clinical depression (Andresen et al., 1994; Irwin, Artin, & Oxman, 1999; Radloff, 1977). The Positive Aspects of Caregiving Scale (Tarlow, Wisniewski, Belle, Rubert, & Ory, 2004); Use of Formal Care and Services Questionnaire (Wisniewski et al., 2003); Self-Ratings of Health Questionnaire (Steffen et al., 2002); and an overall Quality of Life Scale (Logsdon, Gibbons, McCurry, & Teri, 1999) were also administered. Data were also obtained on the care recipients cognitive status (Mini- Mental Status Examination [MMSE] (Folstein & Folstein, 1975)) and functional status (Activities of Daily Living [ADLs] (Katz, Ford, Moskowitz, Jackson, & Jaffe, 1963); Instrumental Activities of Daily Living [IADLs] [Lawton & Brody, 1969]). The Use of Formal Care and Services Questionnaire (Wisniewski et al., 2003) was used to determine how many times per month CGs used the respite service. A program evaluation questionnaire was administered at the 6-month post randomization assessment. All of the measures chosen for the study are widely used in the dementia caregiving literature. Maintenance of the impact of the intervention was evaluated by examining changes in CG outcomes between the 6- and 12-month post randomization assessments. Consent The UHSC Institutional Review Board approved the study. Written consent was obtained from CGs and patients, when possible. Caregivers provided consent on behalf of patients who were unable to do so on their own. The Intervention Community REACH was adapted from the evidence-based REACH II CG intervention program (Belle et al., 2006). As in REACH II, the program was 6 months in duration and included 12, 60-min individual (6 face-to-face and 6 telephone) skill building and educational sessions and 5 telephone support groups. The intervention focused on five areas of CG risk: safety, social support, problem behaviors, depression, and CG health. The program was structured through a protocol that specified the content and activities for each session, which included provision of information, problem-solving and behavioral strategies to manage caregiving problems or care recipient behaviors (e.g., wandering). Each CG received sessions on each of the five areas of CG risk. However, the program was tailored to meet the

4 The Gerontologist, 2018, Vol. 58, No. 2 e133 needs of the CG using a risk appraisal (Czaja et al., 2009). Specifically, in the three sessions that focused on CG strategies to address caregiving problems/care recipient behaviors, the area of emphasis was tailored to the needs of CG. For example, an area of emphasis for one CG may have been managing problems with repeated questioning whereas for another it may have been problems with the care recipient wandering. The five 60-min structured telephone support groups were interspersed with the skill building sessions, and included five to six CGs and a leader. Caregivers also received a Caregiver notebook, which included educational material and a community resource guide. To enhance the feasibility of intervention delivery, the program was modified. Unlike REACH II, which included nine 60-min home-based individual sessions and three individual computerized telephone-based sessions, Community REACH included six 60-min face-to-face sessions and six 60-min telephone sessions. In REACH II after each individual session the interventionists wrote detailed behavioral prescriptions, which were delivered at the next session. In Community REACH, the interventionists were provided with the behavioral strategies developed in REACH II. Further, the assessment battery was shortened. Finally, to integrate the program within existing services provided by UHCS, CGs were also eligible to receive in-home respite for help with household or personal care tasks. Results Analysis Descriptive statistics were used to provide summary information on the CGs and CRs demographic characteristics and the CG outcome variables. Logistic regression was used to examine differences between those enrolled at baseline and those who did not complete the 6-month post randomization assessment and between those who completed the 6-month post randomization assessment and did not complete the 12-month post randomization assessment. We included CG age, gender, depression, quality of life, social support, total bother, use of respite, positive aspects of caregiving, overall burden, strength self-efficacy, respite self-efficacy, disruptive self-efficacy, thoughts self-efficacy, and self-reported health in this models. We did not include bother for depression or memory, to avoid collinearity with total bother. Given that Community Reach was a community implementation of an evidence-based program and did not include a control group, an RCT analytic strategy (e.g., intent to treat) was not employed. We used a series of random effects models to compare baseline and 6-month post randomization assessment and baseline and 12-month post randomization assessment scores and between the 6-month and 12-month post randomization assessments to examine change over time for the outcome measures. In these models, participants were allowed to have a random intercept. The amount of respite used at each time point was the only covariate in the models. This variable was only significantly predictive for the variable use of formal care and services. As expected higher use of amount of respite predicted higher levels of reported use of formal services. Time was entered as a factorial variable. Paired comparisons were performed on each time point. To protect from Type 1 error, contrasts were only interpreted if the total effect of time was statistically significant. Parameters were summarized as means with 95% confidence intervals, p values less than or equal to.05 were considered statistically significant. We conducted also a separate analysis using binomial mixed regression to examine if there was a decline in CGs at risk for clinical depression. Participants One hundred and forty-six CG enrolled in the study who were largely female (76%), providing care for a spouse (45%) or parent (41%). Most of the care recipients (CR) (62%) were female and Hispanic (84%). The CGs average age was 65 years (M = 65.30, SD = 12.94) and the average age of the CRs was 82 years (M = 82.22, SD = 6.19). The average MMSE score of the CRs at baseline was 11.0 (M = 11.09, SD = 7.76). At baseline, 65 of the 146 participants (44.52%), at 6 months 52 of the 112 participants (46.43%), and at 12 months 48 of the 94 participants (51.06%) were spouses of the person they were caring for (see Table 1 for more details on the sample). One hundred and twelve CGs completed the 6-month follow-up and 94 completed the 12-month follow-up. Reasons for drop out between baseline and 6 months included: bereavement (n = 1), placement (n = 3), and unavailability for follow-up (n = 30) and for drop out between 6 and 12 months included: bereavement (n = 11) and placement (n = 7). There were no significant differences between completers and noncompleters at the 6-month post randomization assessment on the set of predictors χ 2 (15) = 17.54, p =.29 entered into the regression models. At the 12-month post randomization assessment, the set of predictors was significant χ 2 (15) = 30.39, p =.01. Specifically, the set of initial use of respite was significant (p =.03) such that CGs who used respite were more likely to stay in the study. Health was also significant (p <.001) such that CGs with lower ratings of their health were more likely to drop out of the study. Outcomes Table 2 presents results of our analyses, including effect sizes, for the comparisons between baseline and the 6-month post randomization assessment; baseline and the 12-month post randomization assessment; and between the 6-month and 12-month post randomization assessment. We also report the mean change from baseline to the 6-month

5 e134 The Gerontologist, 2018, Vol. 58, No. 2 Table 1. Sample Descriptive Statistics Across the Three Assessments Variable Baseline (N = 146) M (SD) 6 months (N = 112) M (SD) 12 months (N = 94) M (SD) CG Age (12.94) (12.68) (11.92) CR Age (6.19) (6.19) (6.30) Years CG 4.28 (4.86) 4.65 (5.18) 4.74 (5.36) Quality of life 2.39 (0.44) 2.43 (0.50) 2.44 (0.42) MMSE CR (7.76) 9.85 (7.74) 7.63 (7.35) IADL CR 7.19 (0.98) 6.91 (1.14) 6.54 (1.44) ADL CR 4.46 (1.77) 4.67 (1.77) 5.00 (1.38) Times using Respite (over the past month) (16.78) (15.75) (15.99) CES-D (7.14) (7.01) 9.64 (7.41) Overall Burden 1.73 (0.82) 1.44 (0.84) 1.28 (0.76) Social Support (7.48) (9.09) (8.60) Overall Bother (15.27) 9.22 (10.52) 7.42 (8.84) Bother-Affective Distress (10.08) 8.93 (9.59) 8.99 (8.65) Bother-Memory 7.64 (6.90) 4.00 (5.30) 2.67 (4.55) Positive Aspects (8.76) (8.51) (9.47) Quality of Sleep 1.28 (.94) 1.71 (1.04) 1.69 (1.06) Strength Self-Efficacy (29.51) (31.05) (29.52) Respite Self-Efficacy (30.00) (35.82) (36.02) Disruptive Self-Efficacy (24.53) (17.92) (17.03) Thoughts Self-Efficacy (24.27) (20.86) (20.25) Note: ADL = Activities of Daily Living; CES-D = Center for Epidemiologic Studies Depression Scale; CG = Caregiver; CR = Care Recipient; IADL = Instrumental Activities of Daily Living; MMSE = Mini-Mental Status Examination. Table 2. Results of Mixed-model Results for Various Dependent Variables Effect sizes for improvement Mean change Dependent Variable F value T1-T2 T1-T3 T2-T3 T1-T2 95% CI T1-T3 95% CI Respite 3.59* ** ( 1.05, 8.89) 6.99 (1.82, 12.15) CES-D ** 0.80 ** 1.04 ** (2.22, 5.25) 5.07 (3.48, 6.67) Overall Burden 7.40 ** 0.39* 0.65 ** (0.03, 0.36) 0.35 (0.17, 0.53) Social Support 6.03 ** ** 0.57 ** 0.08 ( 1.99, 1.84) 3.10 (1.05, 5.15) Quality of Life ( 0.02, 0.20) 0.11 (> 0.01, 0.22) Overall Bother ** 0.73 ** 0.77 ** (3.83, 9.99) 7.87 (4.56, 11.18) Bother-Affective Distress 3.09* 0.44* (0.66, 7.50) 3.50 ( 0.35, 7.34) Bother-Memory ** 0.68 ** 0.95* (1.71, 4.73) 4.83 (3.20, 6.45) Positive Aspects ( 0.51, 4.16) 1.36 ( 1.14, 3.86) Quality of Sleep 5.12 ** 0.41* 0.48 ** (0.08, 0.61) 0.43 (0.14, 0.71) Strength SE 3.96* 0.40* 0.45* (1.04, 13.38) 8.73 (2.10, 15.36) Respite SE ** ** 0.73 ** 3.50 ( 4.22, 11.21) (11.01, 27.49) Disruptive SE ** 0.71 ** 0.82 ** (6.30, 17.49) (8.48, 20.25) Thoughts SE 9.18 ** 0.56 ** 0.62 ** (4.67, 16.68) (6.39, 19.29) Note: Mean change scores are reported for each time interval; positive numbers indicate improvement. Effect sizes are measured with Glass g. *p <.05; **p <.01; T1 = Baseline; T2 = 6 months; T3 = 12 months. Degrees of freedom were estimated with the Satterthwaite approximation; CES-D = Center for Epidemiologic Studies Depression Scale; SE = Self-Efficacy. post randomization assessment; baseline to the 12-month post randomization assessment. Overall, the intervention had beneficial effects. At 6 months, CGs reported significantly less depression (g =.80) (Figure 1), lower overall burden (g =.39) (Figure 2), and lower overall bother (g =.73) with care recipient behaviors (Figure 2). Specifically, they reported less bother with affective distress and memoryrelated problems. They also reported higher self-efficacy with respect to responding to disruptive patient behaviors, and controlling upsetting thoughts about caregiving. There were no significant changes in social support, obtaining respite services, or positive aspects of caregiving (Table 2). The

6 The Gerontologist, 2018, Vol. 58, No. 2 e135 effects for depression, burden, bother, and self-efficacy were maintained at 12 months. At 12 months, CGs also reported higher overall social support and higher self-efficacy obtaining respite. There was a significant increase in use of respite services from baseline to the 12-month follow-up. On average, at baseline, CGs reported using respite services about 22 days a month (M = 22.49, SD = 16.78) and at the 12-month follow-up this increased to about 29 days a month (M = 29.48, SD = 15.99) (Table 2). We found a significant effect of time (F [2, 349] = 6.86, p <.001), such that those at risk for clinical depression showed significant declines in depression at both the 6-month (t [349] = 3.09, p =.002, OR =.55), and the 12-month follow-up (t [349] = 2.95, p =.003, OR =.56) when compared to baseline. There were no differences between the 6-month post-test and 12-month follow up (t [349] =.03, p =.97, OR = 1.01). The percentage of CGs with high levels of depressive symptoms declined from 40% (39.73) at baseline to 21% (21.43) at the 6-month and the 12-month follow-up (21.28) (Table 1). Caregivers Satisfaction and Benefit Of the CGs who completed the evaluation, 96% (N = 71) indicated that they benefitted from the intervention. Figure 1. Means for CES-D at each time. Error bars are 95% confidence intervals. Ninety-three percent (N = 69) indicated that it made their life easier, helped them to better understand memory loss (86%, N = 63); care for their loved one (83%, N = 60); and keep the care recipient living at home (78%, N = 56). Most (61%, N = 43) indicated that the program enhanced the care recipient s life. The CGs also found the online support groups valuable. Specifically, they indicated the groups improved their knowledge of caregiving (83%, N = 44) and their caregiving skills (81%, N = 42). All of the CGs indicated they would recommend the program to others. Discussion This study evaluated the impact of an evidence-based Caregiver Intervention Program, REACH II (Belle et al., 2006) in a community-based service program. Overall, the program was effective. Caregivers who participated in the program reported a significant improvement in depression; burden; less bother with care recipient behavior and memory problems; and greater CG self-efficacy. Participation in the program also resulted in a decrease in the number of CGs at risk for clinical depression. This is an important finding given that at baseline 40% of the CGs had symptoms of clinical depression. These results are similar to those found for the REACH VA program (Nichols et al., 2011), the REACH OUT Program (Burgio et al., 2009), and the REACH II trial (Belle et al., 2006). Importantly, treatment effects were maintained at the 12-month post randomization assessment, which indicates that a relatively shortterm (6 months, 12 sessions) intervention can have lasting effects. REACH VA, REACH OUT, and REACH II did not include a 12-month follow-up assessment. Interestingly, CGs reported an increase in social support and self-efficacy with respect to obtaining respite at the 12-month followup. The increase in self-efficacy in obtaining respite at the 12-month post randomization assessment may be due to the fact that there was also an increase in use of these services. Thus, CGs had more confidence in obtaining and using these services. The increase in social support at the 12 month post randomization assessment may be related to the fact that the CGs learned skills related to asking for support and receiving support in the intervention and support groups sessions and translated the skills they learned Figure 2. (A) Means for burden at each time and (B) Means for Bother at each time. Error bars are 95% confidence intervals.

7 e136 The Gerontologist, 2018, Vol. 58, No. 2 beyond the intervention period to others once the interventionist and formal support groups were no longer available. Caregivers also found the program to be beneficial and indicated that the program increased their ability to provide care, which is consistent with the reported increase in caregiving self-efficacy. They were especially enthusiastic about the support groups, which may in part be due to the fact that they could participate in the groups from the comfort of their own home. Like REACH VA, this study demonstrates that it is feasible to implement evidence-based programs into community settings and that these programs can be effective when delivered by organizational staff. Our sample was different than that included in REACH VA. Our sample was comprised of both spousal and nonspousal CGs and was primarily Hispanic whereas CGs in REACH VA sample were primarily spousal and White/Caucasian (78%). The care recipients in our sample were also considerably more impaired. This provides further evidence for the effectiveness of the REACH II intervention as it indicates that the intervention is efficacious for a broad range of CGs. An important next step would be to examine if a shorter, less intense version of the intervention would be effective. If so, this would enhance the feasibility of broad scale implementation of the program. Nichols and colleagues (Nichols et al., 2016) found that a four-session version of REACH VA had similar outcomes to longer versions of REACH II. Importantly, the community organization was involved in tailoring the intervention so that it could be implemented within the constraints of the organization. The number of sessions delivered in the home was reduced, behavioral strategies were provided, and the assessment battery was shortened. We also integrated the program within the organization s on-going respite program and provided ongoing support for the staff. Having organizational buy-in and input was an essential component of the implementation process (Czaja et al., 2016). This study addresses an important gap in the intervention literature, the implementation of evidence-based interventions in practice. Although there has been an enormous investment in the development and evaluation of CG interventions, relatively few intervention programs have been implemented in clinical and community settings especially with minority CGs. The recent report by the National Academies of Engineering Science and Medicine (2016) on family caregiving, noted that to increase the number of CGs who benefit from proven interventions, these interventions need to undergo a translational phase where the program is adapted to fit within an agency or clinic and evaluated within that setting. Challenges encountered included missing data on some of the assessment instruments such as the program evaluation. This is less likely to occur in RCTs, as there is greater control of study protocols. An additional challenge was participant retention. The care recipients in our sample were fairly impaired; 16% were placed or died over the course of the study. Further, many CGs were unreachable at follow-up; which may have been due to staff having insufficient time and resources to locate these participants. Also, as noted, health of the CG was a significant predictor of drop out at 12 months. Finally, it is difficult as with most community programs, to evaluate the longer-term sustainability of the intervention. Due to changes within the organization and resource constraints, the entire Community REACH intervention is not currently being implemented. However, UHCS staff are still using several components of the intervention, such as the skill building modules and resources guide with their CG clients. The current limited implementation of the program underscores the importance of evaluating the feasibility of interventions with respect to available community resources and constraints. Optimally, this should be done at the front end of intervention development. In order to impact health care practices, understanding how to best design interventions so that they can be implemented in community and clinical settings is an imperative (Gitlin and Czaja, 2015). Funding This study was funded by the Administration on Aging (90AE03204/01). Conflict of Interest None declared. References Alzheimer s Association. (2017) Alzheimer s disease facts and figures. Andresen, E. M., Malmgren, J. A., Carter, W. B., & Patrick, D. L. (1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). American Journal of Preventive Medicine, 10, Barrera, M., Sandler, I., & Ramsay, T. (1981). Preliminary development of a scale of social support: Studies on college students. American Journal of Community Psychology, 9, doi: /bf Bedard, M., Molloy, D. W., Squire, L., Dubouis, S., Lever, J. A., O Donnell, M., (2001). The Zarit burden interview: A new short version and screening version. The Gerontologist, 46, doi: /geront/ Belle, S. H., Burgio, L., Burns, R., Coon, D., Czaja, S. J., Gallagher-Thompson, D., Zhang, S.; Resources for Enhancing Alzheimer s Caregiver Health (REACH) II Investigators. (2006). Enhancing the quality of life of dementia caregivers from different ethnic or racial groups: A randomized, controlled trial. Annals of Internal Medicine, 145, doi: / Burgio, L. D., Collins, I. B., Schmid, B., Wharton, T., McCallum, D., & Decoster, J. (2009). Translating the REACH caregiver intervention for use by area agency on aging personnel: the REACH OUT program. The Gerontologist, 49, doi: / geront/gnp012

8 The Gerontologist, 2018, Vol. 58, No. 2 e137 Capistrant, B. D. (2016). Caregiving for older adults and the caregiver s health: An epidemiological review. Current Epidemiological Reports, 3, doi: /s x Czaja, S.J., Valente, T.W., Nair, S.N., Villamar, J., & Brown, C.H. (2016). Characterizing implementation strategies using a systems engineering survey and interview tool: A comparison across 10 prevention programs for drug abuse and HIV sexual risk behavior. Implementation Science, 11, 70. doi: / s Czaja, S. J., Gitlin, L. N., Schulz, R., Zhang, S., Burgio, L. D., Stevens, A. B., Gallagher-Thompson, D. (2009). Development of the risk appraisal measure: A brief screen to identify risk areas and guide interventions for dementia caregivers. Journal of the American Geriatrics Society, 57, doi: /j x Feldstein, A. C., & Glasgow, R. E. (2008). A practical, robust implementation and sustainability model (PRISM) for integrating research findings into practice. Joint Commission Journal on Quality and Patient Safety, 34, doi: / S (08) Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Minimental state. A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, Gitlin, L. N., & Czaja, S. J. (2015). Behavioral intervention research: Designing, evaluating, and implementing. New York: Springer Publishing Company. Gitlin L. N. & Hodgson N. (in press). Caregivers as therapeutic agents in dementia care: The evidence-base for interventions supporting their role. In: Gaugler, J. & Kane, R. (Eds.), Family caregiving in the new normal. UK: Elsevier, Inc. Gitlin, L. N., Marx, K., Stanley, I. H., & Hodgson, N. (2015). Translating evidence-based dementia caregiving interventions into practice: State-of-the-science and next steps. The Gerontologist, 55, doi: /geront/gnu123 Gitlin, L. N., Reever, K., Dennis, M. P., Mathieu, E., & Hauck, W. W. (2006). Enhancing quality of life of families who use adult day services: Short- and long-term effects of the adult day services plus program. The Gerontologist, 46, doi: / geront/ Irwin, M., Artin, K. H., & Oxman, M. N. (1999). Screening for depression in the older adult: Criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D). Archives of Internal Medicine, 159, doi: / archinte Katz, S., Ford, A. B., Moskowitz, R. W., Jackson, B. A., & Jaffe, M. W. (1963). Studies of illness in the aged. The index of ADL: A standardized measure of biological and psychosocial function. JAMA, 185, doi: / jama Krause, N. (1995). Negative interaction and satisfaction with social support among older adults. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 50, P59 P /geronb/50B.2.P59 Krause, N., & Markides, K. (1990). Measuring social support among older adults. International Journal of Aging & Human Development, 30, doi: / CY26-XCKW-WY1V-VGK3 Lawton, M. P. & Brody, E. M. (1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. The Gerontologist, 9, doi: /geront/9.3_part_1.179 Logsdon, R. G., Gibbons, L. E., McCurry, S. M., & Teri, L. (1999). Quality of life in Alzheimer s disease: Patient and caregiver reports. Journal of Mental Health and Aging, 5, Nichols, L. O., Martindale-Adams, J., Burns, R., Graney, M. J., & Zuber, J. (2011). Translation of a dementia caregiver support program in a health care system REACH VA. Archives of Internal Medicine, 171, doi: /archinternmed Nichols, L. O., Martindale-Adams, J., Burns, R., Zuber, J., & Graney, M. J. (2016). REACH VA: Moving from translation to system implementation. The Gerontologist, 56, doi: / geront/gnu112 Radloff, L. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, doi: / Schulz, R., O Brien, A. T., Bookwala, J., & Fleissner, K. (1995). Psychiatric and physical morbidity effects of dementia caregiving: Prevalence, correlates, and causes. The Gerontologist, 35, doi: /geront/ Schulz, R. Gallagher-Thompson, D., Haley, W., & Czaja, S. J., (2000). Understanding the interventions process: A theoretical/concept framework for intervention approaches to caregiving. In: Schulz, R. (Ed.), Handbook of dementia caregiving (pp ). New York: Springer. Steffen, A. M., McKibbin, C., Zeiss, A. M., Gallagher-Thompson, D., & Bandura, A. (2002). The revised scale for caregiving self-efficacy: Reliability and validity studies. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 57, P74 P86. doi: /geronb/57.1.p74 Tarlow, B. J., Wisniewski, S. R., Belle, S. H., Rubert, M., Ory, M. G., & Gallagher-Thompson, D. (2004). Positive aspects of caregiving: Contributions of the REACH project to the development of new measures for Alzheimer s caregiving. Research on Aging, 26, doi: / Teri, L., Truax, P., Logsdon, R., Uomoto, J., Zarit, S., & Vitaliano, P. P. (1992). Assessment of behavioral problems in dementia: The revised memory and behavior problems checklist. Psychology and Aging, 7, doi: / Teri, L., McCurry, S. M., Logsdon, R., & Gibbons, L. E. (2005). Training community consultants to help family members improve dementia care: A randomized controlled trial. The Gerontologist, 45, doi: /geront/ The National Academies of Sciences, Engineering, Medicine. (2016). aspx?recordid= Wisniewski, S. R., Belle, S. H., Coon, D. W., Marcus, S. M., Ory, M. G., Burgio, L. D., Schulz, R.; REACH Investigators. (2003). The resources for enhancing Alzheimer s Caregiver Health (REACH): Project design and baseline characteristics. Psychology and Aging, 18, doi: / Zarit, S. H., Orr, N. K., & Zarit, J. M. (1985). The hidden victims of Alzheimer s disease: Families under stress. New York: New York University Press.