Family Visits and Symptom Burden in Nursing Home Residents With Cognitive Impairment

Family visits with nursing home residents have been shown to improve residents' quality of life. However, little is known about the association between in-person family visits and physical (e.g., pain) and behavioral symptoms (e.g., agitation) among residents with cognitive impairment. Using data from a large clinical trial of a palliative care program in Maryland and Indiana, we examined the association between family visits and staff-reported resident symptom burden.

Using cross-sectional baseline data from an ongoing multi-state clinical trial involving 194 nursing home residents with cognitive impairment, we conducted bivariate analyses to examine the association between in-person family visits (family involvement) as reported by family members and staff-reported resident symptom burden. Linear regression models assessed the association between symptom burden and family visits in an adjusted framework. The model controlled for other relevant covariates, including resident race, gender, and age, and the care partner's relationship to the resident. Staff-reported symptom burden was measured using a modified version of the Comfort Assessment in Dying End of Life in Dementia (CAD-EOLD) scale. Family involvement was measured by the average weekly frequency of in-person visits reported by the family over the past month.

On average, spouses visited 5 days per week, children 2 days per week, and other family members 1.5 days a week. In bivariate analyses, there was a weak but statistically significant and positive correlation between family involvement and symptom burden (r = 0.16, p = 0.04). However, in the adjusted model, the association was not statistically significant (b = 0.16, p = 0.09).

Bivariate analyses suggest a significant and positive correlation between family visitation and symptom burden. However, this finding was not in the expected direction. Family visitation was associated with residents experiencing greater symptom burden. The association was not statistically significant in the adjusted model.

ClinicalTrials.gov identifier: NCT04520698.