Patient-, caregiver-, and relationship-level characteristics influence heart failure self-care maintenance behaviors in patient-caregiver dyads

Introduction. Many heart failure (HF) patients rely on informal caregivers to assist them with self-care maintenance behaviors that are essential to maintaining clinical stability. However, few studies examining maintenance utilize a dyadic approach that controls for statistical interdependence between patients and caregivers. Hypothesis. Patient-, caregiver-, and relationship-level characteristics influence maintenance behaviors. Methods. This was a secondary analysis of cross-sectional data from an Italian convenience sample of community-dwelling HF patients and their caregivers. Patient and caregiver contributions to HF self-care maintenance were measured using parallel versions of the Self-Care of Heart Failure Index. Patient health status, impairment in activities of daily living (ADL), cognition, and HF-specific quality-of-life (QOL) were measured using the SF-12, Barthel Index, Mini-Mental State Exam, and Minnesota Living with HF Questionnaire, respectively. Clinical HF data (duration and hospitalization numbers) were abstracted from the medical record. Caregiver social support was measured with the Carers of Older People in Europe Index. A single item measure was used to assess dyad relationship quality. Data were analyzed using multilevel modeling (MLM), controlling for dyadic interdependence and common confounders. Results. The average age of the sample (n=364 dyads) was 76.3±10.8 for patients and 57.7±14.6 for caregivers. Most patients were male (57.3%) with NYHA Class II/III HF (72.2%). Most caregivers were female (51.7%) and adult children (53.6%) or spouses (34.0%). Patient and caregiver maintenance were highly correlated (r=0.70), with significant variability across dyads. Patient cognition (β=0.29±0.14, p<0.05), less ADL impairment (β=0.13±0.05, p<0.01), worse physical health (β=-0.28±0.10, p<0.01), longer HF duration (β=0.03±0.02, p<0.05), worse HF emotional QOL (β = 0.47±0.21, p<0.05), number of hospitalizations (β=-1.86±0.89, p<0.05), caregiver gender (female β=3.06±1.4, p<0.05), higher social support (β=0.77±0.31, p<0.05), and better dyad relationship quality (β=2.0±0.75, p<0.05) were significant determinants of patient self-care maintenance. Only longer HF duration (β=0.05±0.02, p<0.01) and worse patient HF emotional QOL (β=1.03±0.23, p<0.001) were significant determinants of caregiver contributions to maintenance. Conclusions. In this analysis of HF patient-caregiver dyads, a combination of patient, caregiver, and dyadic characteristics predicted patient self-care maintenance, while only patient characteristics predicted caregiver contributions to maintenance. This innovative analytic approach provides a robust means to examine the complexity and interdependent nature of dyadic care relationships that influence self-care.