Supporting caregivers of veterans with Alzheimer’s disease and traumatic brain injury: study protocol for a randomized controlled trial

Veterans are included if they (1) have an ADRD or TBI diagnosis at any stage or level of impairment of their illness, (2) receive primary care services from VHA, (3) can identify a study partner who is a family or friend caregiver, (4) do not reside in a nursing home, (5) speak and read English, and (6) have access to a telephone. Caregivers are included if they (1) have been identified as the primary person who provides, or would provide if needed, the majority of assistance to the individual; (2) have plans of providing care at least for 1 year or longer; (3) are available to participate in the intervention protocol with veterans; (4) read and speak English; and (5) have access to a telephone.

Veterans and caregivers are excluded if they either are < 18 years in age, were not contacted within ten phone calls, deny that they have ADRD or a TBI, have difficulty hearing or speaking by telephone, have a terminal illness, have a history of hospitalization for alcohol or drug abuse or severe mental illness (e.g., suicidal tendencies, severe untreated depression or bipolar disorder, or schizophrenia), or are a prisoner or under house arrest. Caregivers also are excluded if they do not consider themselves an informal caregiver (e.g., believe that the veteran is not impaired or does not support them in any capacity) or if they have a serious medical illness limiting their ability to participate (e.g., suicidal tendencies, severe untreated depression or bipolar disorder, or schizophrenia).

The ABC ANSWERS team is structured to maximize the skill sets of specially trained nurses and community health workers who function as care managers (CMs) and CM assistants in a collaborative model with the primary care team to maximize effectiveness in improving QoL and decreasing caregiver burden. The team is supervised by a physician with expertise in caring for patients with cognitive disorders. The intervention team conducts a biopsychosocial needs assessment in person or by phone upon enrollment. This assessment includes a demographic and psychosocial interview focused on achieving problem identification. The program uses standardized assessment tools including the Healthy Aging Brain Care Monitor [37]. The team focuses on problem clarification and reviews the assessment findings, medical records, medication lists, and other associated assessments. If the veteran has unspecified cognitive issues as a diagnosis, the CM also reviews diagnostic testing, brain imaging results, and functional details of the assessment to determine the presence or absence of a likely dementia or TBI diagnosis and to identify reversible and comorbid conditions. The team creates an initial plan and identifies areas needing further assessment after the initial home visit. For complex cases, the registered nurse will assess need for referral for further evaluation with mental health, neurology, or geriatrics professionals. Feedback from the initial assessment and home visit are forwarded to the veteran’s PCP with requested feedback and further direction as needed.

This phase starts after the initial home visit and concludes with the delivery of the nine ANSWERS sessions (Table 1). The goal of this phase is to create an individualized care plan in collaboration with the veteran’s primary care team. After the initial assessment is completed, any urgent medical, behavioral, and/or psychological issues are addressed through consultation with the program ADRD and TBI specialists, as well as with the PCP as needed. During the second visit (in home or by phone), the team reviews the proposed care plan and identified problem list and seeks input from the dyad in prioritizing care issues. From this resulting consensus, the CM discusses the proposed individualized care plan; explains the diagnosis (ADRD or TBI) and natural history; implements appropriate care protocols; reviews, explains, and distributes the corresponding educational handouts for the dyad; and connects veterans and caregivers to eligible services and community resources. Once all urgent needs have been addressed and the care plan is established, the team initiates the first ANSWERS session, shown in Table 1. Each ANSWERS session consists of a 60- to 90-min structured curriculum that provides education, counseling-based skills, and cognitive rehabilitation techniques to veterans and their caregivers for coping with and managing ADRD and TBI. At the end of each session, the ABC ANSWERS team members and the dyad create an action plan, which outlines specific skills and sets of activities that will be practiced until the next session. The action plan states how often a technique should be practiced, defines roles of the veteran and caregiver in practicing each technique, and provides space to document progress and barriers. Sessions 1–6 focus on key topic areas (see Table 1), and sessions 7–9 serve as “booster sessions” that are conducted over the phone to assist with the maintenance, modification, and generalization of skills.

During the follow-up phase, the CM team continues to work with the dyad as outlined by the ANSWERS action plan and based on presenting needs and circumstances. The CM team answers any questions generated from previous visits, collects veteran and caregiver feedback, has the caregiver complete a brief assessment to identify need for specific care protocols, and facilitates caregivers’ participation in an array of community services that are readily available. The CM reconciles medications and reviews medication adherence during both home visits or by phone. Throughout the duration of the follow-up phase, the team continues to work with the dyad and the veteran’s primary care team to monitor, implement, and adjust the individualized care plan as necessary.

The primary outcome is QoL (Quality of Life in Alzheimer’s Disease scale [QOL-AD]) measured with a 13-item scale [39, 40] designed to provide both a patient report and a caregiver report of QoL [39, 40]. Responses are structured in a four-choice format (i.e., poor, fair, good, or excellent) that is consistent across all questions, and all items are rated according to the caregiver’s or veteran’s current QoL. The scale includes appraisal of physical QoL, mood, interpersonal relationships, ability to participate in meaningful activities, financial situation, and overall assessment of QoL. Psychometric properties indicated that the QOL-AD has both content validity and criterion content validity when correlated with the Dementia Quality of Life Instrument (r = 0.69) and the EuroQol EQ-5D scale (r = 0.54). While QOL-AD has good construct validity, both interrater reliability (Cohen’s kappa > 0.70) and internal consistency (Cronbach’s alpha coefficient = 0.82) were excellent [41]. In a more recent evaluation, the QOL-AD demonstrated acceptable reproducibility at 2 weeks with intraclass correlation coefficient (ICC) > 0.80 and internal consistency with Cronbach’s alpha coefficient > 0.70. In fact, Wolak-Thierry et al. [42] reported that the QOL-AD was preferred over the general health dimension of the Duke Health Profile because the QOL-AD was significantly (P < 0.0001) more efficient in time required to complete [41, 42].

The Oberst Caregiving Burden Scale (OCBS) [43] is used to measure (1) caregiving tasks and responsibilities, (2) relationships and interpersonal support, (3) lifestyle, (4) emotional and physical health, and (5) overall personal impact of being a caregiver. OCBS is a reliable and valid questionnaire that rates 15 different caregiving tasks of informal caregivers based on perceived time and difficulty of the task (e.g., providing personal and medical care, assisting with ADRDLs, monitoring symptoms, managing patient’s emotions and behaviors, dealing with finances, and coordinating and seeking health services). Each item is scored on a 5-point response scale. In response to time for each task, the caregiver chooses an amount ranging from no time to a great deal of time spent on the task related to caregiving. In response to difficulty for each task, the caregiver chooses a value ranging from not difficult to extremely difficult. Similarly, there is a subscale regarding level of distress associated with each task.

Subscale scores for perceived time, difficulty, and distress will be obtained by summing across the 15 items scored 1 to 5. Thus, subscale scores will have a range of 15–75, with higher values indicating more perceived time, difficulty, and distress with caregiving tasks. Evidence of reliability, content validity, and construct validity for the OCBS were reported in prior studies of ADRD [44‐46] and TBI [47] family caregivers and people caring for persons receiving cancer treatment [48‐50]. Acceptable internal consistency reliability also has been reported in the context of family caregivers of stroke survivors [51, 52].

We use the 10-item Center for Epidemiologic Studies Depression Scale (CES-D) [53‐55] to determine the impact of the ABC ANSWERS intervention on caregivers’ and veterans’ depression. The CES-D is a 10-item depression scale with a total score ranging from 0 to 30, with higher scores indicating more symptoms of depression and a score ≥ 10 indicating depression [55]. To measure anxiety symptoms, we use the Generalized Anxiety Disorder 7-item scale (GAD-7), which is a 7-item anxiety scale with a total score ranging from 0 to 21 [56]. Both the CES-D and the GAD-7 have good internal consistency and test–retest reliability, as well as convergent, construct, criterion, procedural, and factorial validity for the diagnosis of major depression and general anxiety disorder [56‐58]. We have used both instruments in multiple research studies, including our dementia collaborative care trials [2, 4, 10, 25‐28, 30, 31, 56, 57].

In addition to the primary and secondary measures described above, we also will measure caregiver and veteran satisfaction with life [59‐62]; caregiver’s emotional and physical health strain, feelings of dyadic strain, and role captivity; and the amount of veteran’s distress experienced due to difficulties experienced in completing ADL [63‐66].

Power estimation was conducted using the GLMPOWER procedure in SAS 9.4 software, which takes into account the repeated measures. Power was estimated under two scenarios for potential intervention effects for the primary outcome of veteran QoL. In the first scenario, we assume an early intervention effect with a 0.3 standard deviation (SD) effect size at 3 months and sustained intervention effects of 0.3 SD at 6 and 12 months. In the second scenario, we assume that the effect of intervention is gradual in a dose–response fashion (i.e., 0.17 SD, 0.34 SD, and 0.68 SD effect sizes at 3, 6, and 12 months, respectively). We further assumed a Toeplitz pattern for the correlation structure with correlations of 0.4 for adjacent measures and 0.2 for measures farther apart. With the total enrollment of 200 veteran–caregiver dyads, we expect 10% loss to follow-up rate due to refusal or death. With 180 veteran–caregiver dyads remaining, both scenarios will have > 80% power to detect an intervention effect from a linear mixed model with type I error set at 0.05. Power will be similar for other aims.