An Examination on the Scope of Caregiver Reaction

Presenter: Seiritsu Ogura, Hosei University

Abstract

Background: In the last two decades, the introduction of psychometric instruments, most notably Zarit Caregiver Burden Interview (Zarit 1980), has made it possible to compare their burdens across different cares, cases or countries. Most recently, in applied clinical works, Given et al.'s Caregiver Reaction Assessment (CRA) (Given et al., 1992), a mulit-dimensional instrument, is becoming very popular.

Data: The data were collected through an internet survey of households providing informal cares for their family members. In March, 2008, the authors conducted the third survey on 3802 individuals, out of which 3202 were repeat samples from their last two surveys, and 600 were new random samples. The response rate was 76.2%. For this paper, the authors have used the samples of current primary caregivers, or 1070 families.

Methodology: Zarit Caregiver’s index and CRA scores and their subscales were computed. We have also collected several self-reported measures of their burdens. Taking self-reported measures of burden as the dependent variable, we have regressed it on CRA subscales and direct (self-reported) measurements of caregiver’s daily activities. The null hypothesis is that if CRA subscales capture all the important aspects of care-giving, none of the direct measurements should be significant.

Results: Given the Japanese dataset, the regression analysis suggests that, compared with measures of self-reported burden, CRA does not fully capture the variation in physical and emotional stress of the care-givers, as (1) care grades in LTCI are very significant, (2) actual family expenditure for the long-term care is very significant, (3) time and scope of the care are significant, and (4) problem behaviour dummies are important. On the other hand, adding family relationship dummy variables neither improves the equation’s R2 nor produces any statistically significant coefficients on them, once we control for care grades, implying that the CRA self-esteem subscale does absorb them.