Family Caregiver Strain and Likelihood of Heart Failure Patient Clinical Events: A Replication Study

Abstract

Introduction Higher caregiver strain was identified as a predictor of lower patient clinical event-risk in a previous analysis of Italian heart failure (HF) patient-caregiver dyads. The purpose of this analysis was to determine whether this trade-off in patient and caregiver outcomes was culturally bound, or replicable in a North American sample. Hypothesis We hypothesized that, similar to the Italian study, higher strain in North American caregivers would predict lower patient event-risk. Methods This was a secondary analysis of data from a 3-arm (usual care, patient-family education, and family partnership intervention) randomized controlled trial to improve HF self-care. Data collection occurred pre-intervention and 4 and 8 months post-intervention. Caregiver strain was measured pre-intervention using the 15-item Bakas Caregiving Outcomes Survey (BCOS: range 15-105, higher scores indicate more positive outcomes related to caregiving), and tertiles were used to divide caregivers into low, moderate, or high strain for analysis. Generalized linear regression was used to examine caregiver strain as a predictor of the likelihood of having any event (HF hospitalization, HF emergency room visit, or all-cause mortality) over time, controlling for intervention group and recent HF hospitalization (previous 4 months). Results All patients (n=92) were NYHA Class II or III, with the majority being Class II (n=68, 73.9%). Patients and caregivers were in their mid-fifties on average (56.3±10.1 and 53.4±12.4 years, respectively) and racially diverse (n=58, 63.0% and n=59, 64.8% African American, respectively), with most patients being male (n=57, 62.0%) and most caregivers being female (n=75, 84.2%). The most common dyad type was spousal (n=40, 44.0%). Over the 8 months, 28 patients were hospitalized, 6 had an emergency room visit and 6 died. The average BCOS scores for the low, moderate, and high strain groups were 82.9±12.5 (range 66-105), 61.1±1.6 (range 60-65), and 52.1±7.4 (range 31-59), respectively. There was no difference in events by intervention group. High caregiver strain (as compared to low strain) was associated with 91.9% lower likelihood of the patient having an event over 8 months follow-up (OR=0.081, 95%CI=0.008-0.850, p=0.036). There was no significant difference in event-risk by moderate caregiver strain. Conclusions Similar to previous findings in an Italian cohort, higher caregiver strain predicted lower HF patient clinical event-risk in this North American sample. Patient event-risk reductions may come with a cost for the family caregiver - a cost that has known sequelae (morbidity, mortality) and must be assessed and addressed.