Informal caregivers provide a majority of care for patients at the end of life. Efforts to improve caregiver experience, and thus patient experience, often focus on pain and symptom management and coping skills. However, less is known about effective ways to address other elements of caregivers' role that may increase capacity to care, namely their relationship to the patient.
Two important elements of this relationship experience are end-of-life preparation and completion. Left unattended, caregivers may find themselves less capable of caring for a loved one and unable to make crucial decisions influencing care. We designed the Caregiver Outlook intervention based on the human development literature and robust evidence in health communication and clinical psychology that addresses the value of expressing emotions and stress on health outcomes.
This study will evaluate whether an end-of-life preparation and completion intervention reduces caregiver anxiety, depression, anticipatory grief, and burden, and improves patient quality of life and health care utilization. The project will also analyze the intervention sessions' qualitative content, examining variation associated with gender, ethnicity, social economic status, quality of family communication, and stage of illness, to improve understanding of the intervention mechanisms of change.
Dyads of 143 caregivers and veterans with advanced cancer, congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), or other serious disease complete a battery of pre-test measures and then are randomly assigned into one of two interventions. Caregivers in the "treatment" (or preparation and completion) group will meet with a facilitator three times for a period of forty-five minutes to an hour each, and complete sessions on issues related to relationships 1) life review, 2) regret and forgiveness, and 3) legacy and celebration. Caregivers in the "attention control" group will meet with a facilitator three times for a period of forty minutes each and be asked to complete a relaxation exercise. Pre- and post-test measures will be administered by a blinded interviewer. Measures: Outcomes will be measured via baseline and follow-up interview questionnaires using the FACT-G quality of life scale, the FACIT-sp sub-scale, the Centers for Epidemiologic Study of Depression (CESD) short form, the Profile of Moods (POMS) anxiety sub-scale, the Prolonged Grief Disorder-12 (PG-12) scale, the Caregiver Reaction Assessment (CRA), the Completion and Preparation sub-scales of the Qual-E Fam scale, as well as asking about amount of hours per week spent giving care to the patient.
Caregiver/veteran patient pairs (n=143) completed baselines and were randomized, with 71 in the intervention group and 72 in the relaxation meditation (attention control) group for a total of 286 participants. Caregiver mean age was 60.5 and patient mean age was 66.1. Ninety four percent of caregivers were female and 95% of patients were male. The most common dyad relationship was that of spouse(73%) with the next most common relationships being adult, parent, or sibling. The caregiver sample was 43% African American, 48% Caucasian. It was diverse educationally: 43% of caregivers completed a HS degree or less, 37% with some college, and 20% with a college or graduate degree. Similarly, participants fell across a spectrum of financial security with 11% reporting difficulty paying the bills, 24% reporting needing to cut back on things to pay the bills, 34% with enough for bills with little extra, and 30% with enough to pay the bills and have special things. Twenty-eight percent of caregiver participants reported working full or part time, 59% reported being retired or not working, and 12% on disability. The primary life-limiting illnesses of the patients were: cancer: 66%; chronic obstructive pulmonary disease: 12%; congestive heart disease: 8% end stage renal disease: 8%; and multiple diseases: 5%. At baseline, intervention participants were slightly less educated, on average, and more likely to be caring for non-cancer patients.
Aim 1: Evaluate the impact of the Caregiver Outlook intervention on caregiver anxiety (POMS tension subscale), depression (CESD-10), anticipatory grief, completion, and burden.
Analyses showed no significant difference in change over time for caregivers randomized to the Caregiver Outlook intervention as compared to the relaxation on the following outcomes: caregiver anxiety, depression, completion, preparation, four caregiver burden subscales of esteem, finances, lack family support, and impact on schedule. Caregiver burden's, impact on health subscale differed between the two arms (p =.02) with the intervention reporting greater burden. Responses on the PG-12 suggested only 6% percent with traumatic anticipatory grief, at baseline. We saw no differences in yearning and symptomology between groups over time.
Aim 2: Evaluate the impact of the Caregiver Outlook intervention on patient health service utilization, as measured by number of days at home versus in hospital, emergency department, or nursing facility.
The number of days that a patient used either the VA emergency department or was an inpatient at a VA hospital in the 6 months following randomization did not significantly differ (mean of 4.4 days in both patient groups).
Aim 3: Evaluate the impact of the Caregiver Outlook intervention on quality of life in patients with advanced disease.
There were no significant differences between the Caregiver Outlook intervention and relaxation meditation arms with regard to patient quality of life.
Aim 4: Analyze the intervention session's qualitative content, examining variation associated with gender, ethnicity, social economic status, quality of family communication, and stage of illness to improve understanding of the intervention mechanisms of change.
The qualitative analyses showed expected idiosyncratic responses to questions of relationship review, issues of regret and goals; common themes of caregiving emerged including: Motivation for caregiving, caregiving challenges, learning from caregiving, grief and loss, personal struggle. Distinctions in intervention response content appeared to vary less by ethnicity, relationship and diagnosis and more by individual caregiver's own resources.
The sample includes a large number of caregivers who were not experiencing distress at baseline. Thus, changes in improving distress (anxiety and depression) would not be expected. We were unable to include non-VA utilization in the utilization comparison, and the comparison was for 6 months only.
The trial did not demonstrate differences between the Caregiver Outlook intervention and relaxation meditation on most measures (with the exception of health burden), despite participants self-reported benefit from the intervention. The trial, without a usual care arm, was powered to test differences between comparison conditions. A primary cofounder may be that the relaxation meditation was likely not a neutral condition, but rather brought benefits, as well. Additionally, a majority of caregivers in this sample did not demonstrate distress at baseline, as assessed by levels of anxiety or depression cut-points used in other studies to indicate distress by the POMS anxiety subscale or the CESD-10. On the POMS anxiety scale, those randomized to the intervention averaged 6.9 and those in relaxation 7.3; previous research shows population averages (not anxious is 8.2) thus well below norms for anxiety. Similarly, our sample showed those randomized to the Outlook intervention scored 8.2 on CESD depression and those in relaxation 8.8. The most conservative suggest a score of 10 as indicator of depressive symptomology, other studies suggest at cut-point of 16 In either case, the sample was well below the norm and, on average, did not demonstrate baseline distress; thus, demonstrable change is unlikely.
It is possible that the preparation and completion intervention is not particularly robust for caregivers who are in earlier stages of illness caregiving, or that only a subset of particularly distressed caregivers may benefit from this kind of intervention. Qualitative data suggest the need to identify early caregivers with low emotional and coping resources early in the caregiving trajectory.
Caregiver Outlook was designed to serve as a novel intervention addressing needs not addressed in other caregiver interventions. It was designed to complement to interventions aimed at improving caregiver training in medical and other pain and symptom management needs as well as stress and coping, thus ensuring the VA will have empirically-based interventions to improve comprehensive care. As a manualized intervention, it was proposed to be available to hospice partners as a tool well-suited to address Veteran and Veteran family psycho-social care in serious illness. The intervention did not prove significantly different than the comparison condition of relaxation meditation. The findings guided us about best design for trials in caregivers who are caring for those in serious illness but not hospice eligible. For example, in this trial, caregivers, on average were not experiencing distress (i.e. anxiety or depression) at baseline. As a result, efforts to improve those outcomes would not be possible. Second, it is unlikely that relaxation meditation was a neutral condition and was more likely to be a comparison. A trial including a usual care arm would be necessary to test that. Finally, more attention can be given to screening for caregivers experiencing emotional and existential distress earlier in the course of caregiving. This has informed subsequent work to develop tools in spiritual assessment and screening.
External Links for this Project
Grant Number: I01HX000841-01A1
None at this time.